Abstract
Background
The only classification that exists in neurooncology is the World Health Organization Classification Update of Central Nervous System (CNS) Tumors 2016. However, it only takes morphological and molecular genetic aspects of CNS tumors into consideration. The TNM classification that is important for other tumor types is not useful for CNS tumors. Meanwhile, when we detect signs of CNS tumor on MRI (and other neuroimaging modalities), we focus on the location of the tumor, its spread, signs of dissemination, etc., to decide if tumor removal or biopsy is needed in the given case and to suggest following combination of treatment options. We propose to systematize these data to form a multi-level diagnosis that would base on clinical, anatomical, neuroimaging and morphogenetic findings.
Methods
The diagnosis of a primary CNS tumor should be the following: 1) Primary tumor (first signs) or disease progression. In case of tumor progression type of progression should be specified: local growth (at the site of primary lesion), growth into a new anatomical region (independent of the primary lesion), or dissemination (by brain and/or spinal subarachnoid spread, less often - outside the CNS). 2) Location of the tumor, L (we need to precise structured anatomical regions based on imaging data). Major structure affected: Lb (brain tumor), Ls (spinal tumor).Grading: (a) number of lesions (1 - solitary lesion, 2 - multiple (2 or more) lesions); (b) side affected (left or right). So, we can return the concept of brain gliomatosis that has been excluded from last redaction of the WHO classification of the CNS tumors. 3) Advanced neuroimaging data (T1, T2, T2 FLAIR and so on - so called “radiomics”) including PET / CT with amino acids tracers. 4) Pathological diagnosis based on the World Health Organization Classification Update of Central Nervous System 2016.
Conclusions
Thus, for each case of CNS tumor multi-level diagnosis is formulated, which can be used to determine diagnostic algorithms and therapeutic strategies. Moreover, structured data allows easier understanding of patients’ disease stage for a physician. Possibly, precise classification of tumor clinical and anatomical properties would allow forming more uniform groups for clinical trials. Currently we apply this approach on our database of retrospective and prospective data of glioma patients to evaluate the prognostic significance of suggested clinical, anatomical and neuroimaging modalities in comparison with the prognostic significance of established morphogenetic parameters.