6th Annual Brain Tumor Conference: Pathology

Paul Mishel - pathologist
Neuropathology of brain tumors.
We help establish the diagnosis on biopsies done at UCLA and other institutions.
We work with our colleagues in the tumor board. He examines the tumors under the microscope and comes up with a diagnosis and is also involved in research. We try to figure out how to get the right drug to the right patient. Provide preliminary diagnosis at time of surgery - intraoperative “frozen section”. They identify the type and grade of tumor. We provide molecular information that may help better target patients for treatments. Surgeon removes the tumor, neuropathologist makes intraoperative diagnoses at time of surgery then final diagnosis made later (type & grade), team decides best theraputic approach. What to we look for. Is it a tumor? Does it arise from the brain or a metastasis. Tumor type tells us about the probable cell origin (how it looks and proteins it expresses). Conveys prognostic information and general therapeutic information. The grade tells us how biologically aggressive the tumor is likely to be, grade is based on the presence of certain characteristic microscopic features. Is it a tumor? Come from brain or elsewhere? Cells in primary tumor kind of cannonball shape, primaries have similar cells to the brain itself. Gliomas: pilocytic astrocytomas grade 1 less common, most are astrocytomas 1-4 and oligodendrogliomas, ependymomas, rarely are neuronal tumors, lymphomas, hemangioblastoma (unknown origin). Skull based: meningiomas, pituitary tumors, rare. Gliomas tumors arising from the supporting cells of the brain. Astrocytes - astrocytomas, oligodendrocoytes (cells that sheath nerves of brain - oligodendrogliomas, ependyma- ependymomas. Astrocytomas : indolent (not invasive) but most of them are infiltrating. These are graded on their biological aggressiveness. Most frequent types of glioma, infiltrative nature, spectrum from low grade to high. Grade determined by examining features: high cellularity and cellular atypia, mitotic activity (proliferation rate ) , vascular proliferation, necrosis. Different grades have different age distributions: grade 2 30-40, Oligodendroglioma - another common glioma: rounded cells “fried egg appearance”, patternless sheets, they are graded by features as well similar to gliomas. Meningiomas: usually press on but do not invade brain. Arise from covering of brain (durra), they can be scooped out for the most part. Small subset have a more aggressive appearance. Current grading systems: useful for diagnosis and prognosis but there are shortcomings: morphologically identical tumors may behave very differently, does not yet let us individualize therapy. The Tolstoy metaphor: “All happy families are alike: each unhappy family is unhappy in its own way.” The highly regulated molecular events that are crucial for normal development and function are very similar between individuals. In cancer, genetic and epigenetic alterations lead to highly complex individual tumors. The challenge: find consistencies that can be therapeutically regulated
Traditional approach to cancer diagnosis:
The genomic revolution is transforming clinical medicine. Instead of the current model of population -based risk assessment and empirical treatment. We will move to predictive individualized care based on molecular classification and targeted therapy. Targeted molecular therapy trying to find Achilles heal of cancer. Exploits gene defects and signaling abnormalities specific to cancer cells. Holds the promise of non-toxic treatments. Already being for other types of cancer. The success of this approach will depend on targeting it to the right subset of GBM patients.

Glivec for leukemia, doesn't make you sick but suppresses cancer.
Cancers are a very mixed group, some are susceptible to certain drugs and other not. Targeted inhibitors work in some GBM patients - who are they? Some patients show a clear response. A clinical trial will likely fail to show efficacy if the drug is not targeted to the right patients Tumor should be targeted not on the basis of morphology, but on pathway deregulation. Need to develop ways to identify which patients are likely to benefit. How do go from a patient's biopsy to right drug? Pathway signatures are beginning to be used to identify which individual patient may benefit most from a particular drug. In vivo analysis of signaling pathways may tell us how to individualize therapy for patients. We found a way to examine proteins that are turned on when they should be turned off in GBM patients. We can identify if they are activated in certain patients. EGFR is an attractive target in GBM: frequently amplified, over expressed or mutated. Response limited to a subset: only 10-20% respond to EGFR inhibitors. Molecular determinants of response in GBM unknown. 11/10/05 article in NE Journal of Medicine: “Molecular Determinants of the Response of Glioblastoms to EGFR Kinase Inhibitors” Asked if more of the gene target made person more sensitive to drug, answer was no. There are some proteins that we thought might interact with each other and we found that when both proteins present in tumor cells, patient 50% more likely to respond to Tarceva. (Mark's tumor did not express these two proteins, so Tarceva not a good choice.) In the process of doing clinical testing for these proteins. Understanding why some one doesn't respond is almost as important as trying to figure out why some one does. Malignant glioma patients can benefit from EGFT kinase inhibitor therapy Summary:
- we work as part of the team
- we diagnosis the tumor
- info about the presumed cell of origin its likely life of biological aggressiveness provides a diagnosis and provides basic info about the prognosis and treatment
- we work closely with the neuro-oncologists, surgeons and radiation oncologists to help optimize a treatment plan
- we are at a new and hopeful phase in which we may help to individualize therapy and help determine which therapy is likely to benefit each patient
- our group at UCLA is moving forward with the developing such targeted molecular approaches

Will the pathway signature work in low grade settings? Probably yes, but what we know comes from high grade tumors.

Diagnosis of cancer is based on what it looks like under the microscope. There are some cases where the pathologist can't make a definitive diagnosis because there are not enough cells and trouble getting them without harming the patient.

Cells go about their business through a series of signals. One protein enzymatically interacts with another protein and so on to activate a certain function in a cell. These are called pathways. In cancer, these pathways get distorted so cell can't do what it is intended to do.

Blood brain barrier when normally functioning prevents things from getting into the brain. In brain cancer, the barrier becomes leaky so things can get in. Sometimes can exclude some cancer therapies. Most inhibitors appear to be getting in, in sufficient doses.