Letter in Response to David N. Louis et al , International Society of Neuropathology‐Haarlem Consensus Guidelines for Nervous System Tumor Classification and Grading, Brain Pathology, doi: 10.1111/bpa.12171

I read with great interest the article entitled “International Society of Neuropathology-Haarlem Consensus Guidelines for Nervous System Tumor Classification and Grading” by David N. Louis et al, recently published in Brain Pathology, doi: 10.1111/ bpa.12171. After critical discussion with members of the Vienna brain tumor pathology group, we feel that some of the recommendations are most valuable, whereas others might not yet be considered as guidelines but are rather subject to further discussion. I would like to summarize in brief our view concerning the added value and shortcomings/points of criticism as follows: (A) Added value • The recommendation that tumor entities shall be defined as precisely and objectively as possible in order to optimize interobserver reproducibility, and that room shall be given for descriptive diagnosis including the NOS (not otherwise specified) designation for “gray zone” tumors is important. Increased precision of defined tumor entities will enhance credibility of histological diagnosis among clinicians, and will help to establish biologically uniform groups of tumors, thus providing also an optimized basis for scientific research. It is good to have such a reference in the literature now. • The recommendation that all tissue-based information (histological, molecular) shall be elements of an integrated neuropathological diagnosis is also of key importance. It makes clear that molecular testing is an integral part of the neuropathological diagnosis.As a consequence, and as pointed out by the authors, the inclusion of molecular genetic parameters in the neuropathology report will become a necessity. • The recommendation that molecular information should be incorporated in the definition of some diagnostic entities is relevant and is supported by already existing proof of concept (eg, loss of INI1 protein expression as key diagnostic criterion in the case of atypical teratoid/rhaboid tumor). • The statement that the World Health Organization (WHO) grade reflects the natural history of a given tumor after surgery alone rather than the expected patient prognosis following current therapies is a plausible and pragmatic concept. It is good to have such a clarifying statement as contemporary reference in the literature. • The recommendation to separate certain pediatric entities from their adult counterparts is plausible and justified, considering the age-dependent biological differences and behavior of certain brain tumor types. (B) Shortcomings/points of criticism • The authors state that for each tumor entity, the WHO working groups shall decide whether molecular testing is required or suggested to make the diagnosis. Unfortunately, however, objective criteria (= defined levels of evidence) for tumor marker utility (1, 2) are not further specified or suggested. Lack of such a recommendation will leave room for subjective and heterogenous decisions driven by individual interests, views and composition of the working groups. • The authors suggest a concept of future brain tumor classification, which puts molecular information into a modifier role of histologic classification (see table 5 in their paper). In our point of view, molecular information should not overrule a bona fide histotype. For example, if a bona fide oligodendroglioma is 1p/19q intact, why should the designation of such a tumor be modified to “diffuse glioma (oligodendroglioma phenotype)? In our view, it would make much more sense to use molecular information for separating histologic entities into subtypes, for example, “oligodendroglioma, 1p/19q deleted” and “oligodendroglioma, 1p/19q non-deleted.” This is the generic approach used also, for example, in the hematopathology community. Such an approach is also much more in line with the International Classification of Diseases for Oncology (ICD-O) coding system as it has developed and evolved over decades, and which is at the core of the WHO classification system. • The authors suggest a layered diagnosis format with histologic classification, WHO grade and molecular information listed below the “integrated diagnosis,” which shall be positioned in the top line. We are not aware of any field of tumor pathology where such a format has gained general acceptance in practice, so far. We believe that an integrated diagnosis would better be kept as bottom line, reflecting also the natural sequence in diagnostic work-up of brain tumor cases (step 1: HE; step 2: immunocytochemistry; step 3: molecular testing, step 4: integration of all data). Such a format would correspond to the daily life setting of those neuropathology units where molecular diagnosis is not feasible (eg, for economic reasons in low-income countries). This format would also correspond closer to WHO key requirements (to measure the burden of disease worldwide, which has to rely on conventional histology in countries with low income). • The designation “ISN-Haarlem guidelines” seems somewhat inappropriate, neglecting the fact that some crucial points remain open for continued discussion. Although the authors state in their manuscript that the recommendations are only suggested guidelines, the title and the abstract raise the impression of already fixed guidelines. Further, the ISN label would be much more justified if the authors had waited for the input of dedicated focus group meetings on occasion of the upcoming ISN conference in Rio de Janeiro in September 2014. doi:10.1111/bpa.12187