Wednesday, September 26, 2012

Microcystic meningioma in the frontal lobe of a 57-year-old female

When confronted by giant, pleomorphic nuclei among rather non-meningothelial cells, as depicted below, the neuropathologist given a specimen purported to be a benign meningioma might hesitate before agreeing with the neuroradiologist's and neurosurgeon's presumptive diagnosis.
But in this case, a rather small meningotheliomatous component was present to provide reassurance:
A glimpse of a shattered psammoma body also helped to confirm that this was indeed a grade I meningioma, with a predominantly microcystic component.
The microcystic variant is among the most common of the "nonclassical" meningiomas.

Friday, September 21, 2012

Best Post of the April 2012: Photomicrographs of spinocerebellar ataxia type 6 case

The next in our "Best of the Month" series is from April 26, 2012. It's a follow-up from a prior post in which I showed the gross finding from the same case of spinocerebellar ataxia type 6:

A month ago, I posted gross photographs from an autopsy case of mine of a patient with spinocerebellar ataxia type 6. I am now following up with histologic images for this case. The most striking feature was the marked loss of Purkinje cells, which makes sense because SCA6 involves mutation of a calcium channel that is highly expressed on Purkinje neurons:
Many remaining Purkinje neurons demonstrated degenerative changes, including mis-shapened nuclei
Attenuated cerebellar granule cell layer

More Purkinje cells in bad shape

Cerebellar dentate neurons also dengenerated, but not as extensively as Purkinje cells


Mild involvement of inferior olives. Note the mis-shapen and darkened nuclei of neurons in this location.

Pons minimally effected, with only occasional degenerating neurons as seen in the center of this photomicrograph next to a normal neuron

Wednesday, September 12, 2012

DecisionDx-GBM: Should every glioblastoma patient be getting this test?

What does the neuropathology community think of DecisionDx-GBM? This is a product offered by Castle Biosciences, based in Phoenix, AZ. DecisionDx-GBM is a gene expression profile test developed at The University of Texas M. D. Anderson Cancer Center for the purpose of increasing the accuracy of the prognosis and predicted responsiveness of glioblastoma multiforme to first line radiation plus temozolomide. The test is able to distinguish GBM tumors with a proneural phenotype (tumor signature) from those with a mesenchymal / angiogenic phenotype. Patients with a proneural phenotype tumor who are treated with first line radiation plus temozolomide experience a significantly longer median survival (over 7 years) compared to those patients with a mesenchymal / angiogenic phenotype tumor (approximately 1 year). According to the company website, the assay has been fully validated and has been available for clinical use since 2008. A study is ongoing to determine whether the tumor molecular profile conferring a mesenchymal/angiogenic phenotype is associated with a selective increase in benefit from the addition of bevacizumab to temozolomide and radiotherapy. DecisionDx-GBM is also currently being incorporated into a number of other prospective and retrospective studies.

Is DecisionDx-GBM covered by insurance? Castle Biosciences states that the test receives reimbursement from a number of commercial insurance companies, and appeals to the Administrative Law Judge level for Medicare have resulted in favorable decisions for full payment. Should a patient need to pay out-of-pocket for this test, I could not find on the website what the cost would be.

Should we as neuropathologists recommend the use of this profile? Here is a an example of the kind of report that is generated when one orders the panel of 12 genes (3 of which are for control) upon sending a block of paraffin-embedded fixed tissue to the company. I would be interested in hearing people's opinions regarding this product, and whether there are others on the market which might be comparable. Please post!

Dr. Diamandis develops network to help pathologists interface with AI computational scientists

A neuropathology colleague in Toronto (Dr. Phedias Diamandis) is developing some amazing AI-based tools for pathology and academia. He hel...