‘Major Changes’ in New 2016 WHO CNS Tumors Classification

Four years in the making, the new 2016 World Health Organization (WHO) Classification of Tumors of the Central Nervous System (CNS) includes new genetically identified entities and variants, allowing more diagnostic precision, and a new “layered” approach to diagnosis, an author told participants of the 21st Annual Scientific Meeting of the Society for Neuro-Oncology, held November 17–20 in Scottsdale, Arizona. 

“The changes between 2007 and 2016 are really quite major,” said David N. Louis, MD, Pathologist-in-Chief at Massachusetts General Hospital and the Benjamin Castleman Professor of Pathology at Harvard Medical School.

The 2016 update is the sixth edition of the WHO CNS classification since 1979.

The new classification contains a two-page list of entities-the 408-page tome’s “holy tablets,” said Dr. Louis. “The rest is commentary, so to speak.”

The new classification includes newly recognized entities, variants, and patterns, including IDH (isocitrate dehydrogenase) mutant and wild-type entities and pediatric look-alike entities. The update also includes designation of new, genetically defined entities. “There is now good evidence that these are frequently different beasts and should be considered separately,” noted Dr. Louis. “There are now five subtypes of medulloblastoma.”

It remains a work in progress, Dr. Louis cautioned. “The update did not divine some universal, absolute truth,” he said. “It was the best we could do at the time.”

The new classification does not specify a reporting format because the WHO focuses exclusively on diagnosis. It includes more molecular assay findings criteria, without specifying which methods should be used to obtain them.

“The actual book and how it is used have changed,” Dr. Louis said. “The beginning of each definition section has an italicized clause or phrase-the actual diagnostic criteria that are needed to make the diagnosis. Then classic aspects of the tumor follow.”

Gray breakout boxes provide more detail and discuss problematic issues.

Challenging Process

Promulgating the new classification was a balancing act, Dr. Louis said. Participants sought to address the needs of clinicians and their patients, clinical trialists, population health researchers, policymakers, and health insurers.

“The process was challenging,” he said. The usefulness of clinical data to all of these audiences is “highly dependent on accurate classification.”

The team had to decide how many experts would directly be involved in the process (“democracy or oligarchy,” Dr. Louis quipped), and find consensus on classification precision--whether, when in doubt, to lump or split diagnoses. There was no single, cookie-cutter answer to that question, he noted.

“There are no right answers,” he said. “The answers are different depending on tumor type you’re talking about.”

Panelists also had to assess how widespread availability is for new diagnostic technologies when recommending diagnostic criteria for a document that will be used around the globe.

“There is wonderful cutting-edge technology but if you recommend technology that’s not widely available around the world, is that really going to be useful,” he said.

Several meetings were held, in Haarlem, the Netherlands, and elsewhere, to seek were consensus on the new classification.

“What should you call a histological glioblastoma with an IDH mutation: glioblastoma grade IV, IDH-mutant? Or simply an anaplastic astrocytoma, behaviorally? Or glioblastoma grade III,” Dr. Louis asked, illustrating just one example.

CNS tumor grading is distinct from grading of tumors in other organs, he emphasized. For other tumor sites, a histologic-based name is chosen and then a tumor grade is assigned to individual malignancies, whereas for CNS tumors, the name and grade are “absolutely linked,” he said.

A central question for the update was whether a CNS malignancy can be defined based on histology and genetics, rather than histology alone-and how to accommodate institutions or regions that lack ready access to some diagnostic tools.

At the Haarlem meeting, panelists agreed on a new format of “layered diagnoses” for the updated CNS classification, with a top “integrated diagnosis” pending. Histology, grade, and molecular information are other layers that ultimately inform the integrated diagnosis designation.

“That way, the clinician knows the histology early on but knows there is more coming down the pipe,” he explained. Research suggests this approach allows better prognostic power for survival outcomes than the 2007 WHO Classification, he noted.

That integrated, layered-reports approach will affect pathology practice and will initially increase amendment (error) rates, he warned.

The addition of not otherwise specified (NOS) classifications for anaplastic astrocytomas, oligodendrogliomas, and other malignancies, was “very important,” Dr. Louis said. In real-world clinical practice, one starts with histology, then IDH status, and then looks at other parameters, he said. The NOS categories also accommodated institutions that cannot obtain some molecular diagnostic criteria for their patients.

“NOS is a key concept,” Dr. Louis said. “NOS indicates that there is insufficient information to assign a more specific code. Importantly, these are not specific entities; they designate lesions that cannot be classified into any more precisely defined groups. Ideally, you’d want not to use NOS. But it creates a red flag for subsequent oncologists and pathologists, and creates a ‘wastebasket’ category for future study: we really don’t know what these are; they need to be studied further.”

Controversial Process

The classification updating process was somewhat controversial, Dr. Louis conceded.

“There has been a small group of vocal, research-oriented neuro-oncologists who have been unhappy with the 2016 CNS WHO process,” he said. “Many say the update process was too long between updates and was not transparent and not ‘multidisciplinary.’”

Dr. Louis disagreed with the charge that the process was insufficiently transparent or multidisciplinary.

“The Haarlem meeting received input from more than 150 neuro-oncology, non-neuropathology clinicians,” he noted. A survey sent to SNO members yielded additional comments from more than 350 neuro-oncology clinicians.

The process involved meetings between adult and pediatric neuro-oncologists, neurosurgeons, neuroradiologists, and others, he added.

The question of periodicity or how frequently the classification is updated is more difficult to address, he acknowledged. With growing -omics and data toolkits, the evidence base is growing faster than in the past and needs to be revisited more frequently than in the past.

But shorter periods between updates would be no easy task. Each new updated edition requires its own funding-and online, frequently updated books are “not yet practical,” Dr Louis said. The Consortium to Inform Molecular and Practical Approaches to CNS Tumor Taxonomy-Not Official WHO (cIMPACT-NOW) is, however, in development. Dr. Louis and 17 other experts created the consortium to “provide a forum to evaluate and recommend proposed changes to future classifications” at regular intervals, working between new updated editions of the formal WHO classifications to examine how novel diagnostic tools and data can inform the next update effort.