PD-L1- Prognostic and Predictive State of the Science
Charles G. Drake
-normally is a negative prognostic biomarker in cancers
-as a predictor to the treatment response in melanoma -if you have a PDL1 the treatment could work
-in bladder cancer -PDL1 expression correlates with response to atezolizumab;
- unfortunately, within same cancer, same treatment we get different responses for the 1st line and second line, so different conditions could impact the ability of PDL1 to be a right predictor
Limit Utility
1) spatial Heterogeneity
2) which cells type to quantify
3) competing assay and metrics
In lung cancer is clinically used and recently was announced as a predictor in the blade cancer.
Translational research- in mouse models conflicting results.
Message- still we cannot rely on it as a single biomarker, so in melanoma PDL1 cannot guide the treatment decision.
Tumour Mutational Burden (TMB)
Timothy An-thy Chan
Memorial Sloan Hospital
TMB promising as biomarker to identify those people responding to immunotherapy; as seen bellow still we have a consistent percentage of non responders- unmet medical need
Human T-Cells - two important features to remember:
Tumour is part as non-self for our body (that is helping T-cells to recognize the tumor and attack it)
Immune system has memory, so what was recognize as enemy first time could be identified the second time too.
Having antigens and neo-antigens within tumors is helping the immune system to recognize the tumors as non self.
Multiple cancers benefiting: melanoma, urothelial, lung,, renal etc but we still have negative trials, so..what are the genetic determinants of benefit from immunotherapy (anti CTL4 and anti PD1)?
- mutation and neoantigen burden predict response in lung cancer to anti-PD1
- in melanoma- mutational burden a good indicator of response to ipi (for example)
- melanoma is highly immunogenic - got a number of modified proteins as a consequence of DNA damage (environmental factors)
High TMB (Tumor Mutational Burden) and low TMB - in Checkmate 038, Keynote 012, 028 (solid tumors) (to be checked)
So- Hypermutated tumors - respond to anti-PD1
Clonal neoantigens - predict a better response
Patients with HLA +high TMB - do better on immuno
but
how to integrate different biomarkers is the challenge in the future
Future Directions in Immunotherapy Biomarkers
Kurt Schalper
Challenges in developing biomarkers
foto
rules of cell functionality don't apply for the tumor cell
measure the metrics of T-cells
Tumors tissue biomarkers
CD3, CD8
Low TIL infiltrations
Patients with dormant TIL - DO the best
Active TILs did not so good on immuno (see foto)
By different techniques - we understand the imune compositions -effector memory T cells, T cells activation, candidate targets etc.
Liquid biopsy
foto
TMB could be measured in blood too!
More in 2 abstracts (foto)
Serum IL-level in melanoma
BMS company validate this biomarker
low levels low responses and OS
Circulating immune cells and the response to blockers- good metric in patients with melanoma (foto)
Mieloid Derived Suppressors Cells - could be also measured -predict outcomes in certain cancers.
Integration and implementation in clinical practice - remains a challenge
No one could study alone the multitudes of biomarkers- so collaboration between labs is essential.
Questions:
1)Microbiome -promising but the mechanisms are not yet understood
and How we can we use that clinically?
2)PDL1 and TMB - they are independent predictors, what is the best?
-not yet convincing data what is the better- must be judged in context (first line, second line); and is not actually important what is better because they could be used together
3) Imaging - as biomarkers
4)TMB - could be different but responses to therapies could be sometimes similar -that means that there are also other factors involved (such as tumor microenvironment -hostile conditions for T-cells infiltration).
5)calculating TMB from liquid biopsy -how accurate is?
Related literature -
Manson et al., 2016