Notes on Hypoxia and Immunotherapy Resistance

Introduction to Mike Curran's Work

  • Mike Curran is an associate professor of immunology at MD Anderson in Houston, Texas.
  • He completed his PhD at Stanford and pursued postdoctoral studies with Jim Allison.
  • Curran's work includes seminal research on combinations of immune checkpoint blocking antibodies (anti-PD-1 and anti-CTLA-4).
  • He has been recognized by SITC for his work on costimulatory molecules like anti-4-1BB antibodies.
  • Current research focuses on strategies to address cold tumors (e.g., pancreatic and prostate cancers) by increasing T cell infiltration and responsiveness to immunotherapy.

Overcoming Immune Suppression and Resistance to Immunotherapy

  • Curran's group aims to understand and overcome immune suppression and resistance to immunotherapy.
  • The lecture discusses ongoing work on hypoxia and its effects on immune suppression and resistance to immunotherapy.

Checkpoint Blockade and Cancer Immunotherapy

  • Checkpoint blockade can relieve off-switches on T cells, restoring tumor immunity.
  • Blocking CTLA-4 and PD-1, along with PD-L1 and B7-1, can improve outcomes.
  • Checkpoint blockade was a major innovation recognized with the Nobel Prize.
  • The translation of mouse studies to human patients was a key factor.
  • Metastatic melanoma saw significant survival improvements with CTLA-4 and PD-1 blockade, increasing from 18% to 80% two-year survival rates.
  • This combination became the first FDA-approved checkpoint combination.
  • High levels of immune-related adverse events remain a challenge.

Challenges in Immunotherapy for Cold Cancers

  • The majority of patients have cold cancers with low tumor neoantigen levels and weak inflammatory signatures.
  • These patients respond poorly to existing interventions.
  • In melanoma, 30-40% of patients exhibit innate or acquired resistance.
  • Innate resistance: initial non-responders.
  • Acquired resistance: relapse after initial response.
  • Cold cancers include pancreatic cancer (no response to checkpoint blockade) and prostate cancer (infrequent responses).

Hypoxia and Resistance to PD-1 Blockade

  • Hypoxia is associated with a lack of response to PD-1 blockade.
  • Roger Lowe identified a hypoxia-enhanced gene set signature in innate resistant melanoma patients.
  • Mouse models of prostate cancer show prominent hypoxia, stained with pimonidazole.
  • Tumor hypoxia predicts poor response to various cancer therapies.
  • The immune coldest cancers (pancreatic, prostate) have the greatest depth of hypoxia.

T Cell Infiltration and Hypoxic Areas

  • In normoxic areas of tumors, adequate T cell infiltration is observed.
  • FOXP3 positive Tregs control the antitumor activity of effector cells.
  • Hypoxic areas show a lack of T cells.
  • Without T cells, modern immunotherapy approaches are ineffective.
  • Similar findings observed in lung cancer models.

Metabolic Environment of Tumors and T Cell Function

  • The metabolic environment of tumors can be adverse to T cell function, survival, and proliferation.
  • Hypoxia is linked to multimodal mechanisms of active suppression and nutrient deprivation.
  • Hypoxia is involved in the differentiation of suppressive phenotypes of MDSC and M2 macrophages, which deplete arginine, tryptophan, and cysteine.
  • T cells are sensitive to acidic environments and stop proliferating below pH 6.7 and die below pH 6.5.
  • Lactic acid from the Warburg effect, linked to hypoxia, contributes to a drop in pH.
  • Hypoxia drives upregulation of carbonic anhydrase IX, creating carbonic acid.
  • Adenosine release, also linked to hypoxia, binds to A2A and A2B receptors and is broadly immunosuppressive.
  • Overall, the tumor microenvironment presents a significant challenge to T cell function.

Metabolic Blocks vs. Immune Checkpoint Blockade

  • Immune checkpoint blockade involves one-to-one suppressive interactions (PD-1, CTLA-4, LAG-3).
  • Metabolic blocks are more absolute; T cells cannot function in a pH of 6.3 regardless of their numbers.

Mechanisms of Tumor Hypoxia Suppressing T Cell Function

  • Tumor hypoxia suppresses T cell function through direct and indirect mechanisms.
  • Indirect mechanisms include conditioning of the vasculature to be cytotoxic to T cells and not favoring adhesion, rolling, and extravasation.
  • Also, polarization of immature myeloid cells into suppressor phenotypes and upregulation of adenosine-producing enzymes.

Th302 (Evofosfamide): A Hypoxia-Activated Prodrug

  • Th302 reduces hypoxia in solid tumors, which was discovered serendipitously.
  • In Tramp-C2 tumors, hypoxia is prevalent (35-40% of the volume).
  • Checkpoint blockade alone does little to relieve hypoxia.
  • Th302, alone or with checkpoint blockade, effectively reduces hypoxia.
  • In KPC-derived pancreatic cancer organoids, Th302 attenuates hypoxia over time.

Mechanism of Action of Th302

  • Th302 initiates a tissue remodeling process in hypoxic areas.
  • It kills cells driving abnormal angiogenesis, which are replaced by tissue macrophages (M0 phenotype).
  • M0 macrophages lay down new, healthy, thick vessels.
  • These vessels support better T cell extravasation and infiltration, reoxygenation, and breakdown of hypoxia.

Efficacy of Th302 in Mouse Models

  • Combination checkpoint blockade is effective against Tramp-C2 tumors but more so with Th302.
  • Th302 alone shows curative efficacy dependent on adaptive immunity, which is lost in Rag knockouts.
  • Ifosfamide alone cannot replicate the effect; Th302 requires ablation of hypoxic areas for its unique effect.

T Cell Infiltration with Th302 Treatment

  • Untreated Tramp-C2 tumors have some T cells in normoxic zones but none in hypoxic zones.
  • Th302 alone increases T cell infiltration; T cells extravasate from thicker vessels.
  • Checkpoint blockade increases T cell density in normoxic areas but not in hypoxic zones.
  • The combination of Th302 and checkpoint blockade results in a flood of T cells into normoxic and hypoxic areas.

Reduction of Myeloid Suppressors

  • Myeloid suppressors thrive in hypoxia; HIF-1 and HIF-2 activation favor MDSC differentiation and acquisition of suppressive features (arginase).
  • Th302 diminishes CD11b+GR1+ cells, particularly the GR1 subset.
  • The myeloid compartment is taken over by M0-type macrophages.
  • The combination therapy reduces the whole tumor frequency of MDSCs by about half.
  • Arginase-positive CD11b+GR1+ cells are reduced by about 50%.

Functional Impact on MDSCs

  • The hypoxia that remains in the tumor with treatment does not support MDSC proliferation.
  • MDSCs from treated animals are less suppressive.
  • The combination of immunotherapy and Th302 results in a prolonged effect in the tumor microenvironment.

Conditioning of the Tumor Microenvironment

  • Large tumors were established and treated for one week with Th302 and antibodies, followed by adoptive transfer of congenitally marked bone marrow progenitors.
  • The combination treatment impairs the ability to recruit and polarize bone marrow myeloid progenitors into effective granulocytic MDSCs.
  • Absence of immunotherapy in large tumors allows reestablishment of hypoxia.
  • Maintenance of hypoxia reduction by Th302 requires checkpoint blockade.
  • Combination treatment affects the suppressive phenotype, so MDSCs phenotype, they fail to upregulate arginase leading to an incomplete development, their suppressive phenotype.

Efficacy in Spontaneous Tumor Models

  • Spontaneous TRAMP mice develop carcinoma in situ and local metastatic disease by 16 weeks of age.
  • Treatment started at 16 weeks with Th302 and antibodies for a month, followed by observation until 36 weeks.
  • Untreated mice have tremendous prostate burden (adenocarcinomas and neuroendocrine prostate cancer).
  • Checkpoint blockade alone does nothing; tumors resemble untreated animals.
  • The combination treatment results in normal prostates in about a third of the mice at 36 weeks of age.

Impact on Neuroendocrine Tumors

  • TRAMP mice develop adenocarcinomas (80%) and neuroendocrine prostate cancer (20%).
  • The combination group never developed neuroendocrine tumors, which is statistically striking.
  • The change in the metabolic environment may prevent neuroendocrine development or effectively treat these tumors.

P10 SMAD4 Knockout Prostate Model

  • Significant extension of survival observed in Rhonda Pino's P10 SMAD4 knockout prostate model (pure adenocarcinomas).
  • Genetic mutations lead to fewer passenger mutations.

Changes in the Immune Environment

  • CD8 to Treg ratios improve with the combination of Th302 and checkpoint blockade.
  • A large amount of granulocytic MDSC exist, with five MDSCs for every one CD8 T cell at baseline.
  • The combination therapy reverses this, resulting in five CD8s for every one MDSC.
  • CD4 effector proliferation, CD8 proliferation, and granzyme B production improve with combination therapy.

Antigen-Specific T Cells

  • High expression of 4-1BB enriches for antigen-specific T cells.
  • These cells show a higher activation state, higher interferon gamma production, and higher TNF-alpha production.
  • The metabolic conditioning leads to improved T cell effector function.

Durability of Reconditioning

  • The reconditioning effect persists over a long time (four months after treatment).
  • Combination-treated mice maintain better CD8 to Treg and CD4 effector to Treg ratios.
  • There is a trend toward better CD8 to granulocytic MDSC ratios.

Importance of T Cell Survival

  • T cells must survive to mediate tumor rejection.
  • Reactive CD8s undergo apoptosis in the tumor.
  • If T cells live longer in the tumor, the tumors go away.
  • The combination promotes extended T cell survival.
  • In hypoxic areas, actively dying T cells were detected. But with said therapy the rate of death was substantially less.

Clinical Trial at MD Anderson

  • Advanced cold tumor patients who failed other treatments were treated with evofosfamide and ipilimumab.
  • 18% objective response rate and almost 80% disease control rate were seen.
  • Responses were seen, including metastatic pancreatic patients with stable tumors for over a year.
  • The trial was conservative, with two cycles of EVO upfront and standard three mgs per kg Ipi.

Retreatment Results

  • Patients who had responded well initially were retreated, and nearly all responded upon retreatment.
  • Better than expected, and the responses that were seen that showed a true response by the subjects being treated.
  • When patients show enhanced responses upon retreatment, it highlights the importance of the therapy being administered.

Phase Two Trial Design

  • Longer duration of therapy and added PD-1.
  • Small patient number but very advanced patients showing promise.
  • Responses were profound and quite durable.
  • Advanced prostate cancer patient had about 80% tumor shrinkage, with bone met. It's not clear if it's actually there or not, but his state is in a very good state.

Biomarker Analysis

  • Small trial, but collected PBMC and did see some results statistically significant.
  • Statistically significant increases of CD8 T cells in the PBMC.
  • CD4 T cells also significantly increased in the responders versus nonresponders.
  • Patients saw higher arginase expression tended to do more poorly.

Biomarkers for Hypoxia

  • Assess hypoxia reduction, which looks very promising.
  • Compared signatures via HIF-1 and didn't change/went up versus the responders versus pretreatment.
  • Hypoxia signature went down in the responding patients versus the untreated patients.
  • Genomically responders, therapy seems to reduce the hypoxia, and thus correlates with becoming a responder.

Genetic Signatures

  • Genes that were not responding didn't show in high concentrations of T cells, as the ones that were responding did.
  • Responders had a lot of preexisting innate inflammation. A very strong genetic signal was made to delineate the difference.
  • Responders saw increase in innate inflammation. And could make it over the hump with new tumor regression.

Resistance

  • High oxphos, glycolysis fatty acid, being hypermetabolic and adaptive immune remained disadvantaged with hypoxia's reduction.

Phase 2 Analysis Protocol

  • Adding PD-1, Evofosfamide, which has show promising results.
  • Looking at tumor from PET scans to analyze hypothesis.

Additional Oxphos Inhibitor Trial

  • Looking at HPV vaccine in head and neck cancer for PD-1 to get good outcomes.
  • But resistance comes high OXPHOS.

Hypoxia and Origin Points

  • Consumption via abnormal, tumor angiogenesis, which lead to leaky vasculature and disrupted of O2.
  • As the tumor's upregulate and eat up their resources due to high oxygen consumption, hypoxia tends to rise exponentially.

Studies on Tumor Cells vs Oxygen Metabolism

  • Studied B16 with CTLA, that wasn't affected.
  • They were expected to see gene's and protein up and the process to regulate for tumor factors.
  • CD8 and CD4 expression decrease with decrease in microenvironment activity.
  • And the microenvironment was able to adapt to the stress given hypoxia

Dealing with OXPHOS Concerns

  • IAACS developed an inhibitor for cancer cells, which show a preference in tumors.
  • Combination of all three showed an enhanced and great control over the tumors.
    Had to let the tumor establish longer due to the drug being too overactive.
  • There was proliferation and toxicity mostly CD4 proliferation.
  • Arginase was lost and tumor was upregulated and it was helping drive it.

Closing and Thank You

  • Thanks again to Priya Jayaprakash for doing most of the work to make the presentation.
  • And those with funding being involved to the project as well. Thank you.
  • The research has been very promising overall in what we are trying to show.