### T Cells in Autoimmunity and Tumor Immunity

by Kai Wucherpfennig, M.D., Ph.D.

Chair, Department of Cancer Immunology and Virology, Dana-Farber Cancer Institute
Professor of Neurology, Harvard Medical School

- unique features of MS-associated HLA
- HLA region is a major susceptibility locus for MS (known for a long time)
- how do MHC II genes confer susceptibility?
- genetic map - HLA-DR/HLA-DQ interval
- HLA-DR15 haplotype specifically affected
- first they determined the crystalline structure of HLA-DR15

- MHC Class II protein presents as self-peptide to T cells
- there is a critical area in the binding site of HLA-DR proteins that are important for multiple autoimmune diseases - P4 pocket binding site
- different features in different autoimmune diseases
- beta70 and beta71 residues have different charges
- diverse feature of MHC II system - determines peptide diversity
- ex. in Type 1 diabetes - critical polymorphism at beta57 (P9 pocket) that determines binding strength to HLA-DQ sites

- self-reactive T cells are limited during thymus development
- later, some can escape negative regulation and cause death
- these cause autoimmune disease such as MS
- studying properties of these strange TCR's
- usually peptide antigen is between the helices of the MHC receptor (on the T cell)
- almost all microbe-specific T cells look the same
- self-reactive T cells have a different protein domain structure
- in a self-reactive TCR
- fewer contacts with the antigen, but stronger bonds (compared to anti-microbial TCR binding)
- only specific amino acids in the antigen determine whether the TCR can interact with the antigen
- created fake lipid bilayer with MHC II and adhesion protein
- interactions with TCR's captured using images
- first exposure - form microclusters then start to move around
- self-reactive TCR's moved a lot more than the anti-microbial one
- virus-specific TCR's emit more cytokine signalling
- self-reactive TCR's have lower affinity than virus-specific
- however, it's not because the TCR's dissociate sooner from the antigen (both types stay bound to the antigen for the same amount of time following interaction)
- actually because the self-reactive TCR's don't bind to antigens as quickly

### Clinically Relevant Genomic Alterations in Neurosurgical Oncology

by Ganesh Shankar, MD PhD

Instructor of Neurosurgery, Massachusetts General Hospital

- what determines precision in neurosurgical oncology
- surgery (resection or open biopsy)
- diagnosis
- cytoreduction
- look at
- history
- physical exam
- anatomy
- adjuvant
- chemotherapy
- XRT
- targeted therapy
- look at
- tumor histology
- molecular markers
- genomic alterations
- (he wants to use these in the surgery part of the treatment)
- outcome
- increased PFS/OS
- decrease treatment related morbidity
- role of genomics in oncology
- tumor genetics
- exome
- genome
- transcriptome
- patient w/ craniopharyngioma (benign, locally dangerous)
- vision/headache problem
- two histology versions
- ** (something I missed) \_**
- papillary - BRAF V600E mutation highly recurrent, patient had this mutation
- treatment
- put on BRAF antagonists
- 82% reduction tumor
- this alteration was actually a hotspot that could be detected in the blood
- could mean in the future that you don't need a surgical operation to determine this type of targeted treatment
- looking at NGS for examples
- rhabdoid meningioma - very aggressive, tend to recur, short survival rate but also treatment co-morbidities
- classified as grade 3
- certain hallmark somatic/germline events that segregate various menangiomas
- **Shankar and Santagata Neuro-Oncology 2017 paper on chart with figure on certain mutations for each type of tumor** - look at this paper
- histological features - can't really tell if the meningioma has rhabdoid features
- try to use NGS for this - found they only has mutations correlated with grade 1 meningiomas, put aside data set
- BAP1 tumor suppressor
- involved in cell cycle progression, self-renewal, and differentiation
- somatic mutations in BAP1 have been reported in various cancers
- then did analysis using microarrays on BAP1 expression (on the original data set talked about above) across 322 meningiomas (which were originally determined by IHC)
- the patients who had loss of BAP1 had the aggressive progressors
- some of these patients had a germline CNV w only 1 copy of BAP1 and then lost the other one through a somatic mutation
- spinal cord astrocytomas
- difficult to treat surgically
- rare, usually pediatric
- about a quarter are high grade
- post-op survival ~1 year, usually have decreased neurological function
- have to have big surgeries for essentially just a needle biopsy and then some sort of duraplasty (? - look this up)
- wanted to look for molecular biomarkers for these patients so they can just do a lumbar puncture
- known that:
- grade 2/3 gliomas tend to have highly repetitive alterations in the growth factor pathways (e.g. BRAF) (pediatric), IDH1/2 (adults)
- grade 4, (pediatric) histone, (adult) RAS/p53
- NGS on cohort - found H3F3A mutation for pediatric
- a lot of papers came out demonstrating midline gliomas demonstrated highly recurrent mutations in these H3F3 pathways
- diffuse astrocytomas
- intraop diagnosis of glioma can be challenging
- 10-15% of intraop can't tell what you're dealing with, have to do watchful waiting
- a lot of patients have to have a second resection after the initial biopsy operation
- glioma subtypes are characterized by recurrent molecular genetic alterations
- turns out if you look at just IDH1 and TERT promoter status, you can predict which patients survived better
- made a rapid, quantitative assay they could use intraop
- PACMAN probe (?)
- qPCR method - 35 minutes
- this allowed a second opinion combined with the histology for diagnosis
- how to turn a research assay into a clinically-usable assay
- used for CNS tumors
- any kind of tumors that may have an interest in TERT, BRAF mutations
- don't have to wait weeks for NGS
- want to be able to do surgical resection then inject slow-release agents at the surgical margin while the patient waits for chemo/radiation
- IDH mutations sensitive to certain inhibitors
- NAMPT inhibitors - toxic to the body in levels required to reach the brain
- want to develop ways to lay down particles that elute NAMPT inhibitors over a period of time (microparticle delivery system)
- release drugs up to 7-8 days
- demonstrated NAMPT that they can target IDH mutant cells
- need to know the molecular phenotype of the tumor before using this treatment
- tested in IDH1 orthotoic glioma model - reduction in tumor growth when the microparticles were used against IDH1 human astrocytomas
- increase in survival based on this study