Latest Posts:

Sorry, no posts matched your criteria.

Follow Us:

Back To Top

Finding the true holy grail in immuno-oncology

Article by George Tetley, Founding Analyst in Oncology at DSV

Just over 30 years since the initial experiments, Immuno-oncology has reached the clinic and demonstrated remarkable results and the recognition of a Nobel Prize. The feeling is that this can’t come soon enough given that current mainstay therapies of chemotherapy, radiation and surgery have a high therapeutic burden and often only work in the short term with estimates of over 2/3rds of spending providing zero patient benefit.

Vast amounts of investment have been poured into immuno-oncology ventures, over $9.3bn last year, and even biotech press is almost lavish in its praise. All this gives the impression that we have reached the holy-grail.

Unfortunately, this is far from the case. Immuno-oncology is still largely only suitable for a minority of patients, only available for 1/3rd of cancers and only effective in around 20% of cases. Moreover autologous treatments, where a patient’s own immune cells are used is incredibly difficult to fit into current healthcare regimes, and even allogeneic treatments which use ‘off the shelf’ cells are logistically troublesome.

In our opinion the most impactful approaches are less likely to be found in the proliferation of tweaks to make chimeric antigen receptor (CAR) approaches work, and more likely to involve smarter ways of fully taking into account the integrated complexity of the cancer and immune system. Here we take a brief look at areas where new ventures could create significant impact over the next few years.

 

Immune mediated clearance

Adoptive cell transfer involves the growth of immune cells outside the patient’s body, their activation against cancers and dosing of the patient. Currently two therapies are approved, Yescarta and Kymriah, which both take cytotoxic T-cells, derived from the patient. These are genetically modified to express a CAR which enables them to ‘specifically’ target cancerous blood cells. In practice this is labour intensive, the cells are difficult to grow and there have been some severe reactions against the therapies which cost up to $475,000.

As with many early breakthroughs, CAR-Ts are a blunt instrument. CAR-Ts are matured against an antigen presented on the surface of cancer cells, antigens which are highly heterogeneous within and between cancers and remodel in the face of resistance to treatment. Given that most cancer mutations involve non-surface proteins, it’s clear that any holy-grail will need to target intracellular proteins. There are early players in this field such as Immunocore which uses a T-cell receptor mimic (ImmTAC) to bind intracellular antigens presented on the surface of cells by the HLA complex and recruit T-cell effectors. Whilst this has impressed investors, this strategy is unlikely to represent a broadly applicable solution due to the fact that up to 90% of cancers demonstrate HLA downregulation, leaving ImmTACs little or nothing to bind.

Our immune system is an incredibly complex system consisting of multiple cell types and a huge range of dynamics across receptors and signaling molecules. We are right at the start of understanding how to work with these systems but early work suggests that approaches such as using other cell types such as natural killer (co.) and mesenchymal stromal cells may overcome many of the challenges of sourcing, scaling and patient specificity. Ultimately we’re likely to see a shift deploying multiple cell types in combinations alongside multiple immunomodulatory compounds (see below), to stimulate a more balanced and durable response.  

 

Disrupting the defensive tumour microenvironment

One area of huge progress over the last couple of years has been checkpoint inhibitors. This approach increases the efficacy of immuno-oncology therapies by reducing the immunosuppressant properties of solid-tumor cancers. Tumors mediate this immunosuppressant environment by adapting to highly express PD-1, a receptor that shuts down the cytotoxic T-cell response, and inhibition of this ‘checkpoint’ using a PD-1 binding antibody reverses this effect. The leader in this field is Keytruda, Merck’s latest blockbuster.

Sometimes this works really well, but sadly often not: This is because there are a multitude of other checkpoints and different strategies that tumours use to evade immune activity: Checkpoint inhibition only takes the cancer one step back along a long road to immune evasion. Targeting two checkpoints simultaneously can be beneficial as emerging combination strategies will be, but other factors such as regulatory T cells and macrophages, downregulation of MHC (again) and production of immunosuppressive cytokines all contribute to the resistance of cancers to checkpoint inhibition. What’s needed is approaches that fully consider this as an adaptive system as opposed to point solutions.

Beyond approaches that more fully consider the adaptive complexity of the system, the physical structure around a solid tumour (which are notably difficult to treat) is the key element that maintains this hostile environment. Whilst it sounds like a step back from the elegance of cell based approaches, it might be essential to combine immuno-oncology based approaches with more physical disruption that is capable of temporarily breaking down the extracellular matrix. Either through digestion, chemical modulation or even physical disruption.

 

Maybe cell based therapies aren’t that great after all?

As solid tumours are naturally resistant to immune cell attack, therapies employing cells might in and of themselves be fighting an unnecessary uphill battle. Antibody drug conjugates (ADC), whilst slightly out of favour, could provide similar specificity without the biological unpredictability. The reason for their falling out of favour is due to specificity challenges.

If the very toxic payload is even slightly misdirected to other tissues consequences are serious (e.g. the fatal exfoliation events caused by bivatuzumab mertansine). There is clearly potential for a strategy which employs a triple safety lock, where ADCs are locally delivered, specifically cleaved and specifically active would limit off-targeting and dose limiting toxicities without impinging on efficacy.

Unfortunately, again, most emerging companies are concentrating on one aspect. For instance, ADC therapeutics and ImmunoGen are exploring new payloads, while CytomX looks at ‘probodies’ that only bind to antigens in tumour environments (interestingly Pfizer have recently dumped their collaboration). Whilst this might provide an improvement it is frustrating that there isn’t a more joined up approach to addressing this complexity.

 

The stratification and targeting dichotomy

The major problem with targeting cancers is that they’re heterogeneous both within the cancer and across patients. In an ideal world it would be possible to very specifically target each cancer cell, in practice this is challenging as it requires developing therapeutics (and all the costs associated with that) for smaller and smaller patient groups. Fairly quickly the economics don’t work.

For instance the use bispecific targeting, where two antigens are required for therapeutic activity can decrease off-targeting. However this presents challenges as only a sub-population of the cancer cells will have both these markers and if clearance isn’t swift and complete it is inviting the emergence of resistance.

We expect to see a divergence of strategies in stratification, firstly further work on identifying unified targets and better clinical support tools to quantify the effectiveness of combinations.

In terms of unified targeting we think it’s fair to say that we don’t yet fully understand the generalised characteristics of cancer across all levels from genotype to phenotype expression and there may well be very conserved features that can act as a ubiquitous target. It’s unlikely that there is one magic bullet given the level of heterogeneity but there’s certainly room for better targets beyond the small sample of surface antigens currently in play.    

Finally, as the landscape of therapeutics and patient stratification becomes increasingly complex we expect to see some of the biggest impact coming not from biotech but software that helps clinicians and patients make sense of the options available. Something that we’re looking to solve with ConcR (stealth).

We’re just at the beginning of this paradigm shift, but concentrating on building more general therapies and using them in tailor-made combinations for individuals (as opposed to highly targeted monotherapy) may be the most effective path towards cost-effective and fully efficacious treatment – a true holy grail.

Finding the true holy grail in immuno-oncology

Article by George Tetley, Founding Analyst in Oncology at DSV

Just over 30 years since the initial experiments, Immuno-oncology has reached the clinic and demonstrated remarkable results and the recognition of a Nobel Prize. The feeling is that this can’t come soon enough given that current mainstay therapies of chemotherapy, radiation and surgery have a high therapeutic burden and often only work in the short term with estimates of over 2/3rds of spending providing zero patient benefit.

Vast amounts of investment have been poured into immuno-oncology ventures, over $9.3bn last year, and even biotech press is almost lavish in its praise. All this gives the impression that we have reached the holy-grail.

Unfortunately, this is far from the case. Immuno-oncology is still largely only suitable for a minority of patients, only available for 1/3rd of cancers and only effective in around 20% of cases. Moreover autologous treatments, where a patient’s own immune cells are used is incredibly difficult to fit into current healthcare regimes, and even allogeneic treatments which use ‘off the shelf’ cells are logistically troublesome.

In our opinion the most impactful approaches are less likely to be found in the proliferation of tweaks to make chimeric antigen receptor (CAR) approaches work, and more likely to involve smarter ways of fully taking into account the integrated complexity of the cancer and immune system. Here we take a brief look at areas where new ventures could create significant impact over the next few years.

 

Immune mediated clearance

Adoptive cell transfer involves the growth of immune cells outside the patient’s body, their activation against cancers and dosing of the patient. Currently two therapies are approved, Yescarta and Kymriah, which both take cytotoxic T-cells, derived from the patient. These are genetically modified to express a CAR which enables them to ‘specifically’ target cancerous blood cells. In practice this is labour intensive, the cells are difficult to grow and there have been some severe reactions against the therapies which cost up to $475,000.

As with many early breakthroughs, CAR-Ts are a blunt instrument. CAR-Ts are matured against an antigen presented on the surface of cancer cells, antigens which are highly heterogeneous within and between cancers and remodel in the face of resistance to treatment. Given that most cancer mutations involve non-surface proteins, it’s clear that any holy-grail will need to target intracellular proteins. There are early players in this field such as Immunocore which uses a T-cell receptor mimic (ImmTAC) to bind intracellular antigens presented on the surface of cells by the HLA complex and recruit T-cell effectors. Whilst this has impressed investors, this strategy is unlikely to represent a broadly applicable solution due to the fact that up to 90% of cancers demonstrate HLA downregulation, leaving ImmTACs little or nothing to bind.

Our immune system is an incredibly complex system consisting of multiple cell types and a huge range of dynamics across receptors and signaling molecules. We are right at the start of understanding how to work with these systems but early work suggests that approaches such as using other cell types such as natural killer (co.) and mesenchymal stromal cells may overcome many of the challenges of sourcing, scaling and patient specificity. Ultimately we’re likely to see a shift deploying multiple cell types in combinations alongside multiple immunomodulatory compounds (see below), to stimulate a more balanced and durable response.  

 

Disrupting the defensive tumour microenvironment

One area of huge progress over the last couple of years has been checkpoint inhibitors. This approach increases the efficacy of immuno-oncology therapies by reducing the immunosuppressant properties of solid-tumor cancers. Tumors mediate this immunosuppressant environment by adapting to highly express PD-1, a receptor that shuts down the cytotoxic T-cell response, and inhibition of this ‘checkpoint’ using a PD-1 binding antibody reverses this effect. The leader in this field is Keytruda, Merck’s latest blockbuster.

Sometimes this works really well, but sadly often not: This is because there are a multitude of other checkpoints and different strategies that tumours use to evade immune activity: Checkpoint inhibition only takes the cancer one step back along a long road to immune evasion. Targeting two checkpoints simultaneously can be beneficial as emerging combination strategies will be, but other factors such as regulatory T cells and macrophages, downregulation of MHC (again) and production of immunosuppressive cytokines all contribute to the resistance of cancers to checkpoint inhibition. What’s needed is approaches that fully consider this as an adaptive system as opposed to point solutions.

Beyond approaches that more fully consider the adaptive complexity of the system, the physical structure around a solid tumour (which are notably difficult to treat) is the key element that maintains this hostile environment. Whilst it sounds like a step back from the elegance of cell based approaches, it might be essential to combine immuno-oncology based approaches with more physical disruption that is capable of temporarily breaking down the extracellular matrix. Either through digestion, chemical modulation or even physical disruption.

 

Maybe cell based therapies aren’t that great after all?

As solid tumours are naturally resistant to immune cell attack, therapies employing cells might in and of themselves be fighting an unnecessary uphill battle. Antibody drug conjugates (ADC), whilst slightly out of favour, could provide similar specificity without the biological unpredictability. The reason for their falling out of favour is due to specificity challenges.

If the very toxic payload is even slightly misdirected to other tissues consequences are serious (e.g. the fatal exfoliation events caused by bivatuzumab mertansine). There is clearly potential for a strategy which employs a triple safety lock, where ADCs are locally delivered, specifically cleaved and specifically active would limit off-targeting and dose limiting toxicities without impinging on efficacy.

Unfortunately, again, most emerging companies are concentrating on one aspect. For instance, ADC therapeutics and ImmunoGen are exploring new payloads, while CytomX looks at ‘probodies’ that only bind to antigens in tumour environments (interestingly Pfizer have recently dumped their collaboration). Whilst this might provide an improvement it is frustrating that there isn’t a more joined up approach to addressing this complexity.

 

The stratification and targeting dichotomy

The major problem with targeting cancers is that they’re heterogeneous both within the cancer and across patients. In an ideal world it would be possible to very specifically target each cancer cell, in practice this is challenging as it requires developing therapeutics (and all the costs associated with that) for smaller and smaller patient groups. Fairly quickly the economics don’t work.

For instance the use bispecific targeting, where two antigens are required for therapeutic activity can decrease off-targeting. However this presents challenges as only a sub-population of the cancer cells will have both these markers and if clearance isn’t swift and complete it is inviting the emergence of resistance.

We expect to see a divergence of strategies in stratification, firstly further work on identifying unified targets and better clinical support tools to quantify the effectiveness of combinations.

In terms of unified targeting we think it’s fair to say that we don’t yet fully understand the generalised characteristics of cancer across all levels from genotype to phenotype expression and there may well be very conserved features that can act as a ubiquitous target. It’s unlikely that there is one magic bullet given the level of heterogeneity but there’s certainly room for better targets beyond the small sample of surface antigens currently in play.    

Finally, as the landscape of therapeutics and patient stratification becomes increasingly complex we expect to see some of the biggest impact coming not from biotech but software that helps clinicians and patients make sense of the options available. Something that we’re looking to solve with ConcR (stealth).

We’re just at the beginning of this paradigm shift, but concentrating on building more general therapies and using them in tailor-made combinations for individuals (as opposed to highly targeted monotherapy) may be the most effective path towards cost-effective and fully efficacious treatment – a true holy grail.

About George

George Tetley is looking at the immuno-oncology field for opportunities to start-up companies that will improve outcomes for cancer patients. After really enjoying his research project in immune cell metabolic activation at Trinity College, Dublin he went on to study for a PhD at Cambridge. Over the next four years he discovered and brought forward a peptide active against a commonly dysregulated cancer pathway, which has now progressed into an Astrazeneca collaboration. George is passionate about the translation of scientific research into practice and as such joined DSV, which aims to accelerate this process. Immuno-oncology is booming and is promising revolutions in cancer care, which is something that should get us all excited. Get in touch to learn more via george@dsv.io