AACR18 Update: Lloyd J. Old Award, Overcoming Immunotherapy Resistance, and the Importance of Dendritic Cells

On Day 4 of AACR18, the Sixth Annual AACR-CRI Lloyd J. Old Award in Cancer Immunology was bestowed upon UCLA’s Antoni Ribas, MD, PhD, a leader of the SU2C-CRI Immunology Dream Team and a member of the CRI Clinical Accelerator Leadership.

After receiving the award, Ribas spoke not about his past contributions to cancer immunotherapy that led to the prestigious distinction, but instead looked to the future and walked us through a promising approach that he believes “has the promise to significantly speed up clinical translation of genetically engineered [adoptive T cell immunotherapies].”

In particular, he focused on a strategy that takes hematopoietic stem cells—the cells in the bone marrow that give rise to all other immune cells—and reprograms their genome so that they continuously produce tumor-targeting T cells. Importantly, because a CRISPR/Cas9 genome editing approach was used instead of a viral vector, this approach avoided the potential complications that can result from introducing viruses into the manufacturing process, although the possibility that CRISPR/Cas9 may have off-target effects of its own is still being investigated further.

Already, he’s demonstrated that these engineered T cells are capable of targeting and killing NY-ESO-1-positive cancer cells in mice as well as producing the normal cytokine milieu. Furthermore, he showed that NKTR-214, which stimulates the IL-2 T cell growth pathway, enhanced the expansion, tumor-homing, and polyfunctionality of adoptively transferred T cells.

Overall, Tuesday was a prolific day for Ribas that saw his work featured numerous times. In addition to a plenary talk he gave that morning, which discussed “overcoming resistance to PD-1 blockade,” two later talks—one given by the University of Iowa’s Mohammed Milhem, M.B.B.S., and another by UCLA’s Willy Hugo, PhD—also revealed how Ribas is going about achieving that goal.

Milhem’s talk focused on the combination of PD-1 checkpoint immunotherapy and an immunotherapy that targets the innate immune system—the TLR9 agonist CMP-001, which was administered directly into melanoma tumors. According to Milhem, “the strongest known inducer of IFN production is the TLR9 pathway.” Therefore, given the connection between increased interferon (IFN) activity and responses to PD-1 drugs, Milhem explained that they “thought that adding a TLR9 activator to anti-PD-1 therapy would elicit a response in patients who stopped responding or never responded to PD-1 inhibition,” Milhem explained.

As of March, 85 patients with advanced, PD-1-resistant melanoma had been treated with this approach in a phase 1b clinical trial, and so far 15 of them have seen their tumors shrink, including two complete responses as well as three patients who responded after continuing on the study despite initially progressing. Milhem also noted the striking abscopal effect in these patients, whereby the primary tumor serves as a “vaccine” that stimulates subsequent immune responses against distant tumors, which “is a hallmark of successful intratumoral immunotherapy treatment.”

This trial also involved the University of Pittsburgh’s John Kirkwood, MD, a CRI Clinical Team Leader and member of the CRI Scientific Advisory Council.

The talk by Hugo, who works in the lab of UCLA’s Roger Lo, MD, PhD, a CRI CLIP Investigator from 2014-2016, covered an approach that targeted what he called “innate anti-PD-1 resistance signatures,” or IPRES.

Hugo noted that TGF-beta was the major pathway associated with IPRES, and additionally that PD-1 immunotherapy has been shown to delay resistance to MAPK inhibitors. When treatments targeting all three pathways were combine, it improved responses in mice and enhanced their overall survival through “mechanisms of action … [that] may involve suppression of carcinoma-associated fibroblasts,” according to Hugo.

Following Hugo, Kirkwood discussed two other possible strategies—inhibition of either the JAK/STAT pathway or the phosphoinositide 3-kinase delta (PI3Kδ) pathway—being explored for their ability to enhance the effectiveness of PD-1 immunotherapy.

Preliminary results from a phase Ib trial for solid cancers revealed that while a few PD-1-naïve patients had partial responses to the combination of dual blockade of PD-1 and JAK/STAT (via itacitinib), PI3Kδ inhibiton (via INCB050465) appeared to augment anti-tumor immune activity in conjunction with PD-1 blockade to a greater extent, as evidenced by complete responses in patients with PD-1-naïve lung cancer and bladder cancer. One patient with PD-1 refractory melanoma also experienced complete tumor regression after dual PD-1 and PI3Kδ inhibition.

Later on in Day 4, a session headlined by two CRI scientists highlighted the important role of dendritic cells (DCs) in enabling successful anti-tumor immune responses.

Speaking first was the Massachusetts Institute of Technology’s Stefani Spranger, PhD, a CRI postdoctoral fellow from 2014-2016, who discussed her efforts to identify cancer-promoting pathways whose activity differs between tumors that are infiltrated by “killer” T cells and those that aren’t.

Previously, analysis by Spranger and her postdoctoral sponsor at the University of Chicago, Thomas Gajewski, MD, PhD, a member of the CRI Scientific Advisory Council who has received several CRI grants (including a CLIP grant from 2013-2015), revealed the impact of the Wnt/b-catenin pathway. Speifically, activation of this pathway was associated with melanoma tumors that lacked Baft3+ DCs and failed to recruit tumor-specific T cells. In tumors lacking Baft3+ DCs, not even adoptive cell immunotherapy or vaccination-induced antigen-specific T cell memory were enough to control tumor growth.

This led her to ask, “Why are effector T cells not recruited into the tumor microenvironment?” Attempts to answer that question led her to the discovery that Baft3+ DCs are the predominant source of CXCR3 chemokine ligands. Furthermore, these tumor-residing DCs are required for “killer” T cell recruitment, especially in tumors with low mutational burdens. These trends were reflected by the observation of differences in DC composition between progressing and regressing tumors.

Next was the University of California, San Francisco’s Matthew Krummel, PhD, previously a CRI-funded investigator who now sponsors three CRI postdoctoral fellows—Kevin Barry, PhD, En Cai, PhD, and Megan Ruhland, PhD

In work involving Barry, Ruhland, and Edward Roberts, PhD, another CRI postdoctoral fellow from 2014-2016, Krummel sought answers to several questions related to the activity of DCs, especially those that cross-present antigens, in checkpoint immunotherapy outcomes.

He noted that the presence of intratumotral Flt3-ligand (Flt3L) was associated with higher numbers of cross-presenting DCs and improved overall survival in mice. He then went on to reveal that Natural killer (NK) cells, which he showed were the source of this Flt3L, had “profound interactions” with DCs and helped prolong their survival. Overall, the levels of NK cells and cross-presenting DCs were found to be “tightly linked” and correlated with positive responses to anti-PD-1 checkpoint immunotherapy.

Through several in vivo videos of migratory DCs in action, he demonstrated in impressive detail how they transport tumor antigens to the tumor-draining lymph node, where they share them with other cells. Altogether, he declared, these studies helped clarify a “pattern of antigen handoff that has rules … and is regulated differentially in tumors compared to normal self.”

Finally, Day 4 was capped off with a talk by Memorial Sloan Kettering’s Danny Khalil, MD, PhD, who presented work led by Jedd Wolchok, MD, PhD, an associate director of the CRI Scientific Advisory Council, chair of the CRI Clinical Accelerator Advisory Committee, and leader of two CRI Clinical Accelerator trials.

The studies performed in Wolchok’s lab focused on combination immunotherapy targeting the CD40 and Toll-like receptor 4 (TLR4) pathways. After observing that this combination approach—which involved intratumotral delivery of a TLR4 agonist—induced PD-L1 expression in distant tumors, Khalil revealed that this also restored the sensitivity of mouse melanoma tumors to PD-1 blockade, mediated by the activation of DCs and their accumulation in lymph nodes. These changes were also accompanied by the expansion of tumor-targeting “killer” T cells with enhanced cytolytic capability against tumors.

Khalil also noted evidence of the abscopal effect. This too was dependent on Baft3+ DCs and was lost when the distant tumor was genetically different from the primary tumor.

Be sure to read our final update on AACR18 tomorrow.