Primary cerebral lymphoma: a new classification of the disease aimed at improving the lives of patients

HERNANDEZ Isaïas, doctoral student; Dr. ALENTORN Agusti, Dr. Karima Mokhtari in the team of Pr Khê Hoang-Xuan and Pr Marc Sanson in the genetics and development of brain tumors laboratory at the Pitié-Salpêtrière hospital (ICM)

Primary cerebral nervous system lymphoma (PCNSL) is a particularly curious disease because it originates in lymphatic tissues such as lymph nodes, spleen and thymus, but develops in the brain. In addition, because almost nothing is known about this disease, the options and the effectiveness of diagnosis and treatment are so limited that the consequences are unfortunate for patients and their families. Fortunately, recent technologies allow us to extract, read, decode and interpret the information contained in genetic material. Thanks to a national effort organized by the “Cartes d’Identité des Tumeurs” (CIT) program, we managed to have the largest cohort of patients (with genetic data) in the world which led us to find a new molecular classification PCL tumors and better understand the disease.

What is a PCNSL and the “Cartes d’Identité des Tumeurs” program?

PCNSLs originate from B cells but originate in nervous tissue such as the brain and rarely migrate to other tissues. MRI can suggest the diagnosis but it often needs to be confirmed by tissue biopsy or cerebrospinal fluid (CSF) cytology. Furthermore, current treatments include corticosteroids, chemotherapy and radiotherapy but unfortunately, due to a poor response of the majority of patients to treatment, the median survival is only 26 months.
CIT is a research program dedicated to cancer genomics which has produced a database without equivalent in France and Europe. Genomics is the scientific discipline concerned with the structure and functioning of our genetic heritage. PCNSLs and other cancers result from a particular series of alterations in our genetic heritage, unique to each person, so a treatment may prove effective in one patient and without effect in another. By understanding this diversity, we can improve diagnostic and prognostic tests and also the prediction of patient response to treatments.

How was the new classification of PCNSLs made?

The combination of tumor genomic data, obtained with state-of-the-art technology and associated with patient clinical data (survival, MRIs, etc.) has allowed us to better characterize PCNSL tumors and their molecular signatures. In practice we have used billions of variables and several statistical tools to find tumors that share the most informative and biologically and clinically important variables. This analysis led us to find four different molecular subtypes of PCNSLs linked with crucial implications.

What are the implications of these molecular subtypes of PCNSLs?

Starting with the diagnosis, we have developed an algorithm identifying the molecular subtypes of PCNSLs from simple genomic data. Depending on the prognosis, we found that each subtype responds more or less well to the standard treatment, chemotherapy. For example, the CS4 subtype has a 2.6 times greater response, whereas the opposite happens for the CS3 subtype. Moreover, the CS3 subtype is associated with a specific localization on the brain which could explain this diminished response to conventional treatment. After combining all of our information, we have found that each subtype may respond better to specific treatments, which may allow us to customize options for each patient.

What’s the next step?

Although our research has already allowed us to determine molecular subtypes of PCNSLs and to consider personalized treatments for each, there are still many aspects to explore. We initially plan to carry out experimental tests using cancer cells and animal models to corroborate its feasibility. By remaining bearers of hope, we will continue to walk on the road, the end of which is not far away, in order to improve the lives of patients.

👉 Find the publication here: https://www.annalsofoncology.org/article/S0923-7534(22)04732-9/fulltext

Genetics and mechanisms involved in resistance to brain tumor treatments

In an article to be published in the prestigious journal Nature, Mehdi Touat and Franck Bielle (AP-HP/Sorbonne University) from the “Genetics and Development of Brain Tumors” team of l’Institut du Cerveau (Inserm/CNRS/Sorbonne University) and the neuro-oncology and neuropathology departments of the Pitié-Salpêtrière Hospital AP-HP in collaboration with Yvonne Li, Rameen Beroukhim, Pratiti Bandopadhayay and Keith Ligon of the Dana-Farber Cancer Institute (Harvard Medical School, Boston), highlight genetic changes in certain recurrent gliomas that cause the development of resistance to chemotherapy. In addition to its very comprehensive approach to various aspects of molecular and mechanistic analysis, the study deals with the largest sample ever explored in brain tumours.

 

Gliomas are the most common malignant primary brain tumours in adults. Their treatment is particularly difficult, not only because of their location limiting the extent of surgery, but also because they almost always develop resistance to radiotherapy and chemotherapy treatments. Gliomas thus end up relapsing. However, on an individual level, determining why and how they escape treatment is still very difficult in the clinic.

“There were a few described cases of tumour recurrence with what is called hypermutation, i.e. a tumour with an excessively high number of mutations in the genome compared to the average tumour. The biological phenomena causing this unusual phenomenon and their possible link with the development of resistance to treatment were not known. Furthermore, in other types of cancers, immunotherapy is more often effective in cases of hypermutation, but this approach had not been tested in gliomas. “explains Franck Bielle.

 

This study brought together several expert centres in neuro-oncology: the neuro-oncology team at l’Institut du Cerveau, the departments of neuro-oncology and neuropathology and the Tumorotheque of the Hôpital de la Pitié-Salpêtrière (AP-HP/Sorbonne University/SIRIC CURAMUS), and the departments of neuro-oncology and neuropathology and research teams of the Dana-Farber Cancer Institute, Brigham and Women’s Hospital and Boston Children’s Hospital (Harvard Medical School, Boston). By pooling their resources and integrating public databases and data from a partnership with the company Foundation Medicine, they obtained a global sample of 10,000 tumors, the first of this size for a single cancer study.

“The purpose of this work was to determine the mechanisms of hypermutation in gliomas and characterize their role in resistance to standard treatments. Understanding how the tumour resists allows us to adapt treatments and consider new therapeutic approaches such as immunotherapy. To answer these questions, we have combined for the first time the analysis of a very large base of gliomas characterized by high-throughput sequencing, with experimental approaches in experimental glioma models, and the analysis of data from glioma patients treated with immunotherapy. “continues Mehdi Touat.

Initially, the researchers established the prevalence of the hypermutation phenomenon, which is found in up to 50% of recurrences in glioma subtypes with high chemosensitivity at initial diagnosis. They showed a clear association between the hypermutation phenomenon and temozolomide, the most commonly used chemotherapy to treat gliomas. Hypermutations develop only after exposure to temozolomide treatment and, moreover, if temozolomide was effective on the first tumour.

The second step in the work was to understand the mechanism involved in the development of this resistance. Scientists therefore looked for the presence of specific gene alterations in these hypermutated tumours. They identified 4 genes that were mutated almost systematically, all part of a DNA repair pathway called “Mismatch repair” or MMR. By generating artificial mutations of these genes in experimental models, they demonstrate the development of specific resistance to temozolomide. On the other hand, in vitro, temozolomide applied to cells with inactivated MMR genes produces the same hypermutation as that present in patients.

“We show the clear mechanistic link between temozolomide, the inactivation of these MMR genes conferring resistance to this treatment, and continued exposure to this treatment leading to this very characteristic hypermutation in tumour recurrence. We also show that unlike temozolomide, other treatments used in gliomas remain effective in hypermutation recurrence. “says Franck Bielle.

MMR abnormalities are also frequently found in other types of cancers such as colorectal, endometrial or stomach cancers associated with a strong immune response. In gliomas, these treatment-acquired MMR abnormalities have very specific effects not found in other cancers, including a much weaker immune system response. Brain tumours are believed to have such strong immunosuppressive mechanisms that despite the accumulation of tens of thousands of mutations, they are still not recognized by the immune system as abnormal cells to be destroyed.

“This is a very striking difference from other cancers associated with MMR deficiencies, with a significant therapeutic impact. Indeed, in colorectal cancers with MMR system deficiency, some checkpoint inhibitor immunotherapies have been shown to be highly effective. Clinical data from glioma patients treated with the same approach unfortunately do not show sufficient efficacy for patients. “says Mehdi Touat.

“With these results, we will be able to provide information on the response to chemotherapy at the time of tumour diagnosis and during treatment, particularly in case of recurrence after chemotherapy, where the use of high-throughput DNA sequencing techniques would allow us to tailor treatment in a personalised manner. These discoveries do not call into question the use of temozolomide, which has shown significant gains in survival and does not cause hypermutations in the majority of patients. On the other hand, if this phenomenon is identified, the orientation towards a choice of treatment will be more precise and more effective. The question now is first and foremost how other tumours that do not progress to hypermutation eventually resist chemotherapy. We also need to understand the specificities of the microenvironment of these hypermutated gliomas that prevent the immune system from recognizing the tumor and thus pave the way for immunotherapy in neuro-oncology. “concludes the two clinician-researchers.

 

Source

Mechanisms and therapeutic implications of hypermutation in gliomas, Nature 2020.

Brain Tumours: Developing Individualized Therapies

Functional consequences of HDI1 and CIC mutations on the cells of origin of oligodendrogliomas

Oligodendrogliomas are primary tumours of the central nervous system with morphological characteristics of oligodendrocytes. It has been proposed that the precursors of oligodendrocytes or OPC, for Oligodendrocyte Progenitor Cells, are the cells that cause oligodendrogliomas. They are generated from neural stem cells during development. At birth, the latter differentiate into oligodendrocytes, the cells responsible for the myelination of neurons. OPC persist in the adult brain where they are the most proliferative cell type.

 

The genetics of oligodendrogliomas in humans is well known and shows frequent mutations in several genes. For example, the mutation of the enzyme Isocitrate Dehydrogenase 1 (IDH1R132H) is found in 100% of oligodendrogliomas. In physiological condition, IDH1 allows the conversion of isocitrate to a-cetoglutarate, but when mutated, it acquires a new function: that of reducing a-cetoglutarate to D-2-hydroxyglutarate, an oncometabolite that accumulates in cells. Similarly, all oligodendrogliomas have a co-deletion of chromosomes 1p and 19q. It is interesting to note that the 19q arm codes for the CIC (Capicua) protein which is found mutated in 50-70% of oligodendrogliomas. Physiologically, CIC decreases cell proliferation and promotes cell differentiation; thus its mutation in oligodendrogliomas results in a non-functional protein.

 

New therapies are needed for oligodendrogliomas, especially for the more aggressive forms that can progress to glioblastoma.  Although the mutations present in human tumours are well characterized, the research community does not have a study model for understanding the formation and development of oligodendrogliomas. Possessing in vivo models to study the initiation, formation and progression of oligodendrogliomas remains a major challenge in order to discover functional treatments.

 

Emmanuelle Huillard’s research team proposes to create and characterize a new in vivo model to induce HDI1 and CIC mutations in OPC. Our working hypothesis is that the synergy of these mutations will induce an increase in proliferation to the detriment of OPC differentiation, two behaviours that are modified during tumorigenesis.

 

By developing and describing this model, which summarizes the genetics of human oligodendrogliomas, we hope to provide the scientific community with a biological tool for understanding the formation of these tumors as well as for preclinical studies.

 

Sandra Joppe,

post-PhD, Emmanuelle Huillard’s research team

 

“A glioma to kill “

Short film by Nathalie Magne, post-PhD of the Sanson-Huillard’s team who, on the occasion of the Pariscience festival held at the end of October, won first prize in the short film competition in 48 hours with her project “un gliome à abattre”.

More informations : https://pariscience.fr/symbiose-48h/

https://icm-institute.org/fr/actualite/nathalie-magne-post-doctorante-a-licm-remporte-competition-de-courts-metrages-scientifiques-symbiose/

Chordoid glioma: a subtype of brain tumour still poorly understood

Chordoid gliomas are rare brain tumours, resulting from the excessive and cancerous proliferation of specific cells composing the neuronal microenvironment.

It develops exclusively in the third anterior ventricle, the brain cavity close to the base of the skull (circled part on the MRI on the right).

It is most often manifested in middle-aged women (30-60) by memory loss and headaches. It is non-invasive (no metastasis) but has a poor prognosis because of its location and surgical morbidity.

The pathological, histological and molecular characteristics of this tumour are still relatively unknown since only about 85 cases have been reported since its description in 1998 (Brat et al).  The low number of cases and the rarity of samples, due to the difficulty of surgical removal of the tumours, are the main obstacles to further study of this type of glioma.

The teams of Professor Marc Sanson and Dr Emmanuelle Huillard have succeeded in collecting 16 cases of chordoid gliomas and using current sequencing techniques, they have deciphered the genome of these tumours and highlighted a mutation of a gene present in all the samples. The altered protein present in the samples is well known since it is altered in many cancers. However, this mutation has never yet been observed in other types of cancers, making it a possible signature of chordoid gliomas.

A detailed investigation of the molecular consequences of this “founder” mutation is underway in order to identify the mechanisms of development of these tumors and to establish new therapeutic perspectives for this rare subtype of glioma.

 

A recurrent point mutation in PRKCA is a hallmark of Chordoid Gliomas.
Rosenberg S, Simeonova I, Bielle F, Verreault M, Bance B, Le Roux I, Daniau M, Nadaradjane A, Gleize V, Paris S, Marie Y, Giry M, Polivka M, Figarella-Branger D, Aubriot-Lorton MH, Villa C, Vasiljevic A, Lechapt-Zalcman E, Kalamarides M, Sharif A, Mokhtari KM, Pagnotta SM, Iavarone A, Lasorella A, Huillard E, Sanson M.
Nature Communications, 2018, sous presse.

The SonoCloud First project, led by Prof. Alexandre Carpentier, head of the neurosurgery department at Pitié-Salpêtrière and coordinator of the GRC Neuromachine Interface, aims to offer a new therapeutic response to the treatment of glioblastoma. The current median survival of glioblastoma is 15 months, with 1st line treatments combining surgery, radiology and chemotherapy whose efficacy is limited by the blood-brain barrier. The team has developed an ultrasound-emitting medical device (SonoCloud), implanted during tumour removal surgery to prevent recurrence. It allows transient permeabilization of the blood-brain barrier during chemotherapy sessions, thus increasing the degree of treatment delivery.  The SonoCloud First project, coordinated by AP-HP, will receive funding of 2.9 million euros.

Visit the SonoFirst website : https://sonofirst-trial.eu/

More information about Carthera : https://carthera.eu/fr/sono-cloud-2/for-a-selective-brain-therapy/