Fall 2006

Dear Friends,

We are writing to update you on our progress over
the last three
months. First, we have now shared our pilocytic
astrocytoma genetic data
(some generated at Johns Hopkins and some by Dr.
Cowell at Roswell Park)
with Dr. David Gutmann (Washington University in
Saint Louis) in order
to pool a greater number of tumors and find
important genes. This will
take some time, but widening our collaborative
effort is an important
step.

Second, we are just returning from a meeting
sponsored by the World
Health Organization (WHO), at which 30
neuropathologists from around the
globe met for two days in order to update the
brain tumor classification
scheme used as the basis for all diagnosis,
grading and treatment by
neuropathologists, neurposurgeons and
neuro-oncologists. The last such
meeting occurred six years ago.

You will be happy to hear that pilomyxoid
astrocytomas will be
recognized as a distinct tumor variant in the new
scheme, and classified
as grade II (standard pilocytic astrocytomas are
grade I). Recognition
of the pilomyxoid astrocytoma as a distinct (and
often more aggressive)
tumor type by the WHO should help doctors
identify it more easily and
test potentially more effective therapies. This
WHO group decision was
based to a large degree on the work of Dr.
Burger, and we are continuing
our attempts to microscopically define this tumor
more precisely.
Finally, we heard at this meeting that several
other groups are making
progress on unraveling the genetics of pilocytic
astrocytomas, and we
hope to help follow up on some of their findings.

Sincerely.

Charles Eberhart and Peter Burger

Summer 2006

Dear Supporters of Pilocytic and Pilomyxoid Astrocytoma research. We would like to briefly update you on our progress this Spring and Summer. We would like to highlight two areas of progress. First, we just got back gene _expression data from 15 pilocytic tumors. (As you recall, this was one of the major aims of our initial proposal). The studies done previously with Dr. Cowell looked at what DNA was gained and lost in the tumors. This new data addresses how RNA is affected. (To briefly review, DNA makes RNA which then makes proteins). We used Affymetrix gene chips, with over 30,000 genes represented on each chip, to look at RNA _expression in the 15 tumors. We are now getting ready to look at additional tumors using this technique. We are using this data two ways. First, we have compared the changes in RNA _expression to the _expression pattern in normal brain. By doing this, we have identified a group of genes that are altered in tumors, and which are potential therapeutic targets. Second, we are going to compare the changes in RNA with the gains and losses in DNA, and use the combination of the two techniques to hone in even more on the most important genes altered in tumors.

The second advance we would like to mention concerns the growth of pilocytic tumors in culture. As we have mentioned before, one of the challenges that has slowed research into pilocytic astrocytomas is that they did not grow well in culture using standard techniques. However, we and others are developing new ways to culture tumors that may solve this problem. Shown to the right are cells from a pilocytic astrocytoma removed at Johns Hopkins which we have been growing in culture for several months now as “spheres”. The blue blobs are the nuclei of each cell, the green lines are their “cytoplasm” and the orange stain marks several cells that are more primitive and may be the cells needed to keep the tumor growing. We are very excited about this new approach, as with some modifications and improvements we hope it will give us a new way to test drugs on pilocytic tumors.

Finally, we would like to inform you that Dr. Eberhart has secured some additional funding support for research on pilocytic/pilomyxoid tumors affecting the optic nerve from the Johns Hopkins Ophthomology Department which he will use to help move this work forward. We hope your children are doing well, and thank you once again for your wonderful support!

Best Regards,

Charles Eberhart and Peter Burger

Early 2006

Dear Supporters of Pilocytic/Pilomyxoid Astrocytoma Research,
We are writing to update you about the progress we have made since our last report in May.
Pathological review of Pilomyxoid tumors. Dr. Burger and I have organized and reviewed over 63 cases with at least some pilomyxoid features. Of these, 28 were “classic” pilomyxoid tumors, while 21 had both pilomyxoid and pilocytic regions, a category we call “transitional”. A final group of 14 had pilomyxoid features that were not sufficiently well-developed to include in either group. Based on this analysis, and on other cases in which tumors that were initially pilomyxoid recurred as pilocytic astrocytoma, we are increasingly of the opinion that rather than representing a totally distinct tumor type, pilomyxoid astrocytomas are related to the more conventional pilocytic ones. We consider this finding to be a positive one, as it suggests that pilomyxoid tumors can, under some circumstances, be induced to turn into a more indolent type of lesion. We plan to revise and extend this study, incorporating more cases and some clinical data into the analysis, and to write and publish a paper suggesting the two tumors are linked. Dr. Burger has also recently extended the genetic spectrum in which pilomyxoid tumors are observed by publishing a case in the setting of neurofibromatosis (Pediatr Blood Cancer. 2005). Finally, we are still planning to support Dr. Ken Cohen in forming a pilomyxoid astrocytoma tumor registry with which to track outcomes and effective treatment approaches.
Immunohistochemical analysis of Pilomyxoid tumors. Our immunohistochemical study of pilomyxoid tumors has been slowed by the fact that relatively little tissue is resected and available for study in many of these cases. We are working to collect as many tissue blocks as possible, but it is taking some time. Our initial analysis of a limited set of cases suggest that individual cells in the pilomyxoid tumors express both glial and neuronal markers, as well as nestin, a marker of neural stem cells. The fact that the tumors have many immunohistochemical properties of the “stem cells” that form the normal brain is interesting and may have therapeutic implications (see below).
Notch signaling in piloid tumors. Evidence of Notch pathway activation has recently been found in some pilocytic tumors. This is of great interest to us, as we have shown that Notch is active in stem-like tumor cells, and are working with several drug companies to develop Notch-based therapies for brain tumors. The stem-like immunohistochemical profile of pilomyxoid tumor cells we describe above would be consistent with this as well. We therefore plan to confirm the finding that Notch is active in pilocytic and pilomyxoid tumors using immunohistochemistry and quantitative RT PCR. If we do document Notch activity in these tumors, we will see if the anti-Notch drugs we are working with can inhibit pilocytic tumor growth in model systems. As I noted in my last report, no established pilocytic cell lines exist. Such lines are needed if we are to test new therapies in the laboratory. We are therefore trying to establish low-passage lines, as well as sublines immortalized with c-Myc or hTERT. We are also exploring the possibility of collaborating on this with outside labs.
Analysis of DNA gains and losses in Pilocytic tumors. This is being performed in collaboration with an expert in this technique, Dr. John Cowell at Roswell Park Cancer Institute. We sent DNA from 28 pilocytic astrocytomas last fall to his laboratory. He used a DNA “array” to determine which chromosomal areas are gained or lost in the tumors. Over half of the samples have been successfully processed, generating a number of DNA regions that appear to be altered in the tumors. I have inserted below a picture showing data from chromosome 8 in one of the pilocytic astrocytomas. The dots above the line at 1 are DNA gains, while dots below this line are DNA losses. In the tumor shown, two regions of gain are seen in the middle, while one region of loss is shown on the right. The next steps are to (1) identify regions of DNA which are affected in multiple tumors, which suggests they are generally important in pilocytic astrocytomas, and (2) confirm the finding independently using another technique. We are currently planning these steps, and are also enlarging the initial “array” analysis to include more cases.
RNA __Expression Profiling. We are now ready to begin our RNA __expression profiling experiments. As I mentioned in our last letter, these are quite expensive, costing up to $1,000 per sample analyzed. We therefore wanted to wait on doing this until we were sure we had developed the best strategy. We have decided to link this part of the study (at least initially) to the DNA analysis shown above. Recall that DNA is what you inherit from your parents, and it serves as a template for making RNA. RNA in turn serves as a template for making proteins, which make up our bodies. If regions of DNA are gained in tumors, we would expect important genes in that region to have increases in their RNA __expression. Similarly, if DNA is lost, genes that are lost will generate less (or no) RNA. In the diagram of DNA gains and losses shown above, each dot represents a region of DNA that can contain 10-100 genes. We need to figure out which of those genes in the regions gained and lost are the ones important in the formation and growth of the tumor. By comparing how RNA from each of the candidate genes goes up or down and in regions where DNA is commonly altered we hope to indentify candidates. After we have indentified important genes and pathways, the next step will be designing therapies that target those genes and the proteins they encode. The RNA analysis will be performed in our lab here at Johns Hopkins, and the data will then be analyzed in collaboration with Dr. Cowell at Roswell Park.
How money is being spent. To date, we have spent approximately $6500 on the studies described above. Now that we have accumulated reagents and preliminary results, and feel the funds you have worked so hard to raise can be spent to the greatest effect, we have begun to phase in our more expensive basic science investigations. We are currently spending (or planning to spend soon) money on the collection of additional DNA and RNA, on immunohistochemical analyses, on real-time RT PCR and Affymetrix analysis of RNA, and on the development of low-passage cultures. While it is difficult to know the precise cost of these varied lines of research, we estimate that $50,000-$75,000 will be spent over the next six months. If we successfully identify new molecular targets for therapeutic intervention, it will certainly be expensive to develop new treatments that take advantage of this. We therefore encourage you not to halt your ongoing fund raising efforts. We will also attempt to raise more resources to fund the work on our side by applying for grants from the NIH and from private brain tumor foundations.
Thank you once again for partnering with us in these investigations. We hope this report is not too technical, and are happy to answer any questions that you have. We are very excited by the developments to date, and will continue to give you updates as the research moves forward. We are focused as much as possible on investigating therapeutically relevant questions, and hope that our work will eventually generate new treatments that can be used in the future for patients with piloid brain tumors.
Sincerely,
Charles Eberhart, MD, PhD
and Peter Burger, MD

2005
Dear Supporters of Pilocytic and Pilomyxoid
Astrocytoma research. We would like to update you on
our progress this winter. Our pathological review
of pilocytic and pilomyxoid specimens continues,
as does our collaborative effort with Dr. Cowell
to identify changes in the DNA of these tumors.
We would like to briefly describe two additional
projects: creation of a registry for pilomyxoid
tumors, and analyses of Notch signaling in these
tumors.

Pilomyxoid Tumor Registry
Because pilomyxoid tumors are rare, it is
important to gather information from cases around the
country (and from other countries) in order to
accumulate enough data to draw conclusions about
demographic trends, how the tumors look under the
microscope, and how they respond to various types of
therapy. Hopefully, over time this will allow us
to determine which treatments are most effective
in pilomyxoid lesions, and if pilomyxoid and
pilocytic astrocytomas are truly distinct entities.
To accomplish this, we are constructing a central
registry for pilomyxoid astrocytomas. This effort
will involve Drs. Burger and Eberhart in
pathology, and Dr. Kenneth Cohen, a pediatric oncologist
who has treated several patients with pilomyxoid
astrocytomas. If clinicians from other
institutions want to become involved as well, we are
certainly open to extended collaborations. In order to
create a registry such as this, it is necessary
to obtain regulatory approval from the local
institution, and we are currently pursuing this at
Johns Hopkins. We have also met together several
times in order to decide what demographic and
clinical details should be included in the database.
The data you have already collected at
http://health.groups.yahoo.com/group/pilomyxoid,
together with the additional cases we have
assembled at Johns Hopkins will make a very good start.
For a variety of reasons, we plan to have
patients and their families submit the information about
when the tumor was diagnosed, what sort of
treatment has been given, and how each child is doing.
We hope to have you help us create this form, and
plan to send you an early draft sometime in the
next month so you can help us identify and modify
problematic questions.

We are also designing a website that describes
our research, and refers interested parties to a
page that details how to submit materials to the
registry. On that page will be a downloadable
consent form, as well as the questionnaire to be
completed and returned by email, Fax, or regular mail.
Copies of medical records could also be sent in
at this time if they are available. Over time we
may move to a full web-based data entry system.

A critical aspect to the registry is central
pathological review. It is not always clear if a
tumor if a pilocytic astrocytoma, pilomyxoid
astrocytoma, or a mixture of the two. Many pathologists
have not had extensive experience with pilomyxoid
lesions, and we feel it will be important for
experts to verify the diagnosis of cases in the
registry. Drs. Burger and Eberhart will therefore
review the microscopic slides for each case in the
registry to determine to what extent it is
pilomyxoid. This will be done free of charge, and a
formal consultation report will be generated for each
case. In many cases, local institutions will be
able to send slides to us directly for this
"second opinion" upon the families/patients request. In
some instances it will be necessary for families
to get the slides from the local hospital and
mail them themselves. Patricia Goldthwaite, a data
coordinator with our group, will maintain the
database and contact families each year for updates.
Please feel free to give us feedback at any point
on these plans!

Notch Signaling in Pilocytic and Pilomyxoid
Astrocytoma
In prior updates I had mentioned that we planned
to investigate the possibility that the Notch
signaling pathway might be "turned on" in pilocytic
and pilomyxoid tumors. Notch signaling plays a
critical role in the development of the brain in
the womb and during childhood. We have already
shown it is also important in several other types of
brain cancer. One reason we are interested in the
pathway is that drugs that can turn it off have
already been developed as possible treatments for
Alzheimer's disease, and we hope to use these in
the future to treat brain tumors. Before we test
such drugs in pilocytic/pilomyxoid tumors, we
need to show that Notch is active in them. To do
this, we measured the levels of a gene called Hes1
that is activated by Notch. The most sensitive and
specific way to do this is using tumor tissue
that is frozen shortly after being removed from the
patient. Our initial study indicated that Notch
signaling is active in most of the 20 pilocytic
astrocytomas we examined. This is good news, as it
means that we can move forward with testing
anti-Notch drugs against pilocytic tumors grown in
culture. As we have mentioned before, few good
culture model systems exist for testing drugs against
pilocytic astrocytomas. We will therefore have to
either generate such systems ourselves, or
collaborate with the few laboratories in the country
that have developed such systems. We are pursuing
both these approaches, but it will take time.

You will notice that no pilomyxoid tumors are
present in the experimental group - this is because
they are rare, and sometimes difficult to remove
in large pieces. We can only freeze tissue for
research in cases where we have a lot of tumor to
start with, because first priority must be given
to microscopic examination of the tumor to
establish a diagnosis. We are making the accumulation of
frozen material from pilomyxoid tumors a
priority, but this will also take time. We are therefore
pursuing other ways of testing Notch activity in
pilomyxoid tumors that can be performed using
microscopic slides like those used for the initial
diagnosis.

In summary, we are making slow but steady
progress on several fronts. We are excited about the
upcoming tumor registry, and by the results
suggesting anti-Notch drugs may ultimately be used in
these tumors. While it unfortunately takes time, we
believe that these are very promising first steps
that will be relevant to improving the lives of
children with pilocytic and pilomyxoid tumors.
Thank you once again for your support that has made
these studies possible.

Best Regards,

Charles Eberhart and Peter Burger

PS - Please feel free to forward this email and
any prior reports to other interested parties.

Charles G. Eberhart MD PhD
Assistant Professor in Pathology
Johns Hopkins University School of Medicine
720 Rutland Ave, Ross Bldg 558
Baltimore, MD 21205
Tel. 410-502-5185
FAX 410-955-9777
ceberha@jhmi.edu