Lillian Siu (Princess Margaret Cancer Centre, Canada)
“Immune and Other Biomarkers of Resistance to Anti-PD1 Therapy in NPC”
PDL1 in NPC, Limitation of ORR as a surrogate of benefit with IO therapies Other buiomarkers of interest (HLA-I, TMA, immunoprofuiling, radio is signatures, microbiome)
HSC et al. PD-L1 >= 1% tumor cells or TILS (22C3) J Clin Oncol 2017
Ma et al. J Clin Oncol 20128
Hamid et EMSO 2017
Lessons from Spacadostat (IDOi) + Pemnbrolizuman ASCO 2018 Ling et al
ORR might not be the appropriate point for clinical trial
HSC et al 2017
ORR =26%
1 yr OPFS = 33%
1yr OS = 66%
Ma teal J Clin Oncol 2018
+20% tumor increase.
+- PD-L1 is an enrichment biomarker but nor a perfect predictive biomarker in NPC
Need to look for other poteintial predictive bimarkers
Importantly, need to be able to select out patients other than ID-L1/PD1
Management of Cnacer int he post-anti-PD-1/L1 era
Fo rImmune desert or Immune excluded Tring T cells into tumors. (Generat/expand T cells)
–
Maximal heterozygosity at HLA-I loci (A, B, C_ improved overalll survival after immune checkpoint blockade in melanoma compared to homozugosiuty in at least 1 choices. (Crowell et al, Science 2017) <— still open issue
Correlation between tumor mutational burden and objective response rate awith anti-PD-1 or anti-PD-L1 therapy in 27 tumor types (Yarchoan et al NEJM 2017). <— got to look it up
Tumor mutational burden (TMB). Ali et al Cancer 2017
NPC type (TMB > 10 per MB)
NP-sseuqmaout cell cancer N =62 (15%)
Foundation medicine (undifferentiated NPC only 5 % TMB)
SFL: might be worthwhile to look at the paper, talking about EBV-associated gastric cancer versus Gastric cancer
KW Lo: the specimens in Foundation Medicne might need to involve non-real NPC cases, he suggests those samples should be further validated for EBER or pathology examination
——
INSPIRE (investigator initiated phase II study of pembrolizumab Immunological response evaluation. <—. Clothier et al, Submitted.
– Immune profiling them Church of BEn WEang and Pamela Ohasshi
– 4-1BB+ PS-1 double-Poitier CD8 T cells at baseline could be a poteintial biomarker for an it-PD1 therapy
Radionics
– hypothesizes that medical images, considered as highly multivariate data sets, contain quantitative patterns byeyound those discernible to the naked eye
– Marries the extraction of texture featyurew with machine learning to build quantiutiative predictive
Sun et ADCO 2018 French group C, Ferile. Soon will show up in Lancet Oncol
Medical image cimpiting to assess tumor infiltrating CD8 T cells, tumor immediately pjemtupe
Radionics wirkfliw and prediction of OS by radio I UC
Prairie at al ESMO 2017
Evaluation of a predictive radio,is signature for response to immediate checkpoint inhibitor
Radionics data from keynote-012 to cross validate PM radio,ic signature
(Not start yet)
——-
Microbiome and immunotherapy. (The human microbiome – the collection of microorganism, that live on and in us, their genomes and products, is essential for human health and may affect response to immunotherapy
MIcrobiome influences response to immunotherapy
Science 2017,
– gut microbiome influences efficacy of PD-1 naked immunotherapy against eputhelial;I’ll tumors
– Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma…
ROMA project : role of microbiome as a biomarker in locally-advanced OPC (LA-OPC)
PCoA plot of Braun-Curtis distances <— pretty cluster plot
Microbiome subdues in CCTG HN.9