Álvaro Padrón

Tumor-intrinsic PD-L1 signals regulate cell growth, pathogenesis and autophagy in ovarian cancer and melanoma.

PD-L1 antibodies produce efficacious clinical responses in diverse human cancers, but the basis for their effects remains unclear, leaving a gap in the understanding of how to rationally leverage therapeutic activity. PD-L1 is widely expressed in tumor cells, but its contributions to tumor pathogenicity are incompletely understood. In this study, we evaluated the hypothesis that PD-L1 exerts tumor cell-intrinsic signals that are critical for pathogenesis. Using RNAi methodology, we attenuated PD-L1 in the murine ovarian cell line ID8agg and the melanoma cell line B16 (termed PD-L1lo cells), which express basal PD-L1. We observed that PD-L1lo cells proliferated more weakly than control cells in vitro As expected, PD-L1lo cells formed tumors in immunocompetent mice relatively more slowly, but unexpectedly, they also formed tumors more slowly in immunodeficient NSG mice. RNA sequencing analysis identified a number of genes involved in autophagy and mTOR signaling that were affected by PD-L1 expression. In support of a functional role, PD-L1 attenuation augmented autophagy and blunted the ability of autophagy inhibitors to limit proliferation in vitro and in vivo in NSG mice. PD-L1 attenuation also reduced mTORC1 activity and augmented the antiproliferative effects of the mTORC1 inhibitor rapamycin. PD-L1lo cells were also relatively deficient in metastasis to the lung, and we found that anti-PD-L1 administration could block tumor cell growth and metastasis in NSG mice. This therapeutic effect was observed with B16 cells but not ID8agg cells, illustrating tumor- or compartmental-specific effects in the therapeutic setting. Overall, our findings extend understanding of PD-L1 functions, illustrate nonimmune effects of anti-PD-L1 immunotherapy, and suggest broader uses for PD-L1 as a biomarker for assessing cancer therapeutic responses. Cancer Res; 76(23); 6964-74. ©2016 AACR.

Immunotherapy for Ovarian Cancer

Opinion statementAll work referenced herein relates to treatment of epithelial ovarian carcinomas, as their treatment differs from ovarian germ cell cancers and other rare ovarian cancers, the treatments of which are addressed elsewhere. Fallopian tube cancers and primary peritoneal adenocarcinomatosis are also generally treated as epithelial ovarian cancers. The standard of care initial treatment of advanced stage epithelial ovarian cancer is optimal debulking surgery as feasible plus chemotherapy with a platinum plus a taxane agent. If this front-line approach fails, as it too often the case, several FDA-approved agents are available for salvage therapy. However, because no second-line therapy for advanced-stage epithelial ovarian cancer is typically curative, we prefer referral to clinical trials as logistically feasible, even if it means referring patients outside our system. Immune therapy has a sound theoretical basis for treating carcinomas generally, and for treating ovarian cancer in particular. Advances in understanding the immunopathogenic basis of ovarian cancer, and the immunopathologic basis for prior failures of immunotherapy for it and other carcinomas promises to afford novel treatment approaches with potential for significant efficacy, and reduced toxicities compared with cytotoxic agents. Thus, referral to early phase immunotherapy trials for ovarian cancer patients that fail conventional treatment merits consideration.

Chronic mTOR inhibition in mice with rapamycin alters T, B, myeloid, and innate lymphoid cells and gut flora and prolongs life of immune-deficient mice

The mammalian (mechanistic) target of rapamycin (mTOR) regulates critical immune processes that remain incompletely defined. Interest in mTOR inhibitor drugs is heightened by recent demonstrations that the mTOR inhibitor rapamycin extends lifespan and healthspan in mice. Rapamycin or related analogues (rapalogues) also mitigate age‐related debilities including increasing antigen‐specific immunity, improving vaccine responses in elderly humans, and treating cancers and autoimmunity, suggesting important new clinical applications. Nonetheless, immune toxicity concerns for long‐term mTOR inhibition, particularly immunosuppression, persist. Although mTOR is pivotal to fundamental, important immune pathways, little is reported on immune effects of mTOR inhibition in lifespan or healthspan extension, or with chronic mTOR inhibitor use. We comprehensively analyzed immune effects of rapamycin as used in lifespan extension studies. Gene expression profiling found many and novel changes in genes affecting differentiation, function, homeostasis, exhaustion, cell death, and inflammation in distinct T‐ and B‐lymphocyte and myeloid cell subpopulations. Immune functions relevant to aging and inflammation, and to cancer and infections, and innate lymphoid cell effects were validated in vitro and in vivo. Rapamycin markedly prolonged lifespan and healthspan in cancer‐ and infection‐prone mice supporting disease mitigation as a mechanism for mTOR suppression‐mediated longevity extension. It modestly altered gut metagenomes, and some metagenomic effects were linked to immune outcomes. Our data show novel mTOR inhibitor immune effects meriting further studies in relation to longevity and healthspan extension.

Tumor cell-intrinsic PD-L1 promotes tumor-initiating cell generation and functions in melanoma and ovarian cancer

As tumor PD-L1 provides signals to anti-tumor PD-1+ T cells that blunt their functions, αPD-1 and αPD-L1 antibodies have been developed as anti-cancer immunotherapies based on interrupting this signaling axis. However, tumor cell-intrinsic PD-L1 signals also regulate immune-independent tumor cell proliferation and mTOR signals, among other important effects. Tumor-initiating cells (TICs) generate carcinomas, resist treatments and promote relapse. We show here that in murine B16 melanoma and ID8agg ovarian carcinoma cells, TICs express more PD-L1 versus non-TICs. Silencing PD-L1 in B16 and ID8agg cells by shRNA (‘PD-L1lo’) reduced TIC numbers, the canonical TIC genes nanog and pou5f1 (oct4), and functions as assessed by tumorosphere development, immune-dependent and immune-independent tumorigenesis, and serial transplantability in vivo. Strikingly, tumor PD-L1 sensitized TIC to interferon-γ and rapamycin in vitro. Cell-intrinsic PD-L1 similarly drove functional TIC generation, canonical TIC gene expression and sensitivity to interferon-γ and rapamycin in human ES2 ovarian cancer cells. Thus, tumor-intrinsic PD-L1 signals promote TIC generation and virulence, possibly by promoting canonical TIC gene expression, suggesting that PD-L1 has novel signaling effects on cancer pathogenesis and treatment responses.

Considerations for successful cancer immunotherapy in aged hosts

ABSTRACT Improvements in understanding cancer immunopathogenesis have now led to unprecedented successes in immunotherapy to treat numerous cancers. Although aging is the most important risk factor for cancer, most pre‐clinical cancer immunotherapy studies are undertaken in young hosts. This review covers age‐related immune changes as they affect cancer immune surveillance, immunopathogenesis and immune therapy responses. Declining T cell function with age can impede efficacy of age‐related cancer immunotherapies, but examples of successful approaches to breach this barrier have been reported. It is further recognized now that immune functions with age do not simply decline, but that they change in potentially detrimental ways. For example, detrimental immune cell populations can become predominant during aging (notably pro‐inflammatory cells), the prevalence or function of suppressive cells can increase (notably myeloid derived suppressor cells), drugs can have age‐specific effects on immune cells, and attributes of the aged microenvironment can impede or subvert immunity. Key advances in these and related areas will be reviewed as they pertain to cancer immunotherapy in the aged, and areas requiring additional study and some speculations on future research directions will be addressed. We prefer the term Age Related Immune Dysfunction (ARID) as most encompassing the totality of age‐associated immune changes. HIGHLIGHTSThis article reviews major findings in age effects on immunity and how that affects cancer immunotherapy efficacy.Differences and similarities between mouse models and human data are discussed.This background is used to summarize key findings in efforts to use cancer immunotherapy successfully and optimally in aged patients.Major reviews are addressed and original source data are discussed.Areas for further investigations and testing are discussed.

Biphasic rapamycin effects in lymphoma and carcinoma treatment

mTOR drives tumor growth but also supports T-cell function, rendering the applications of mTOR inhibitors complex especially in T-cell malignancies. Here, we studied the effects of the mTOR inhibitor rapamycin in mouse EL4 T-cell lymphoma. Typical pharmacologic rapamycin (1-8 mg/kg) significantly reduced tumor burden via direct suppression of tumor cell proliferation and improved survival in EL4 challenge independent of antitumor immunity. Denileukin diftitox (DD)-mediated depletion of regulatory T cells significantly slowed EL4 growth in vivo in a T-cell-dependent fashion. However, typical rapamycin inhibited T-cell activation and tumor infiltration in vivo and failed to boost DD treatment effects. Low-dose (LD) rapamycin (75 μg/kg) increased potentially beneficial CD44hiCD62L+ CD8+ central memory T cells in EL4 challenge, but without clinical benefit. LD rapamycin significantly enhanced DD treatment efficacy, but DD plus LD rapamycin treatment effects were independent of antitumor immunity. Instead, rapamycin upregulated EL4 IL2 receptor in vitro and in vivo, facilitating direct DD tumor cell killing. LD rapamycin augmented DD efficacy against B16 melanoma and a human B-cell lymphoma, but not against human Jurkat T-cell lymphoma or ID8agg ovarian cancer cells. Treatment effects correlated with IL2R expression, but mechanisms in some tumors were not fully defined. Overall, our data define a distinct, biphasic mechanisms of action of mTOR inhibition at doses that are clinically exploitable, including in T-cell lymphomas. Cancer Res; 77(2); 520-31. ©2016 AACR.

Age effects of distinct immune checkpoint blockade treatments in a mouse melanoma model

&NA; Cancer immunotherapy has shown remarkable recent progress. Immune checkpoint blocking antibodies have become the most successful anti‐cancer agent class ever developed, with six distinct agents approved since 2011 for a wide variety of cancers. Although age is the biggest risk factor for cancer (aside from selected early‐onset pediatric cancers), these agents were tested pre‐clinically in young hosts, and there is remarkably little published on the effects of host age on treatment outcomes in pre‐clinical studies or human clinical trials. The three principal immune checkpoints against which blocking antibodies have been FDA‐approved for human use are CTLA‐4, PD‐1 and PD‐L1. We used a mouse model of transplantable, orthotopic B16 melanoma to test age effects of treatments with anti‐CTLA‐4, anti‐PD‐1 and anti‐PD‐L1 antibodies. All three agents were highly effective in treating young tumor‐bearing hosts as expected. Anti‐PD‐L1 as a single agent had no effect on tumor growth in aged hosts, anti‐CTLA‐4 had detectable, modest effects and anti‐PD‐1 was essentially as effective in aged as in young hosts, the first single agent we have identified not to lose efficacy with age in this model. Other important differences in young versus aged hosts included lack of anti‐CTLA‐4‐mediated depletion of intratumor regulatory T cells in aged hosts and poorer ability of all three agents to activate T cells in aged versus young hosts. Anti‐CTLA‐4 efficacy appeared to improve when combined with anti‐PD‐L1. Regulatory T cell depletion with FDA‐approved denileukin diftitox did not improve treatment by any single agent. Aged mice tolerated treatments as well as young mice without obvious toxicities at equivalent doses. HighlightsDemonstration that &agr;PD‐1 alone can be as effective in aged mice with B16 melanoma as young miceSingle agent &agr;PD‐L1 or &agr;CTLA‐4 are poorly effective in aged mice with B16 melanoma, despite good efficacy in young hosts.&agr;CTLA‐4 + &agr;PD‐L1 has survival impact in a young subset with B16, indicating that the same could be possible in aged mice.Regulatory T cell depletion (denileukin diftitox) does not improve immune checkpoint blockade effects in aged mice with B16

Abstract 5649: Cell-intrinsic PD-L1 directs tumor initiating cell fate and rapamycin and interferon-γ response

Tumor initiating cells (TIC) foster treatment resistance and tumor relapse. We show here that in mouse B16 (melanoma) and ID8agg (ovarian carcinoma), and human ovarian cancer ES2, PD-L1 knockdown (shRNA) reduced TIC numbers and functions. Data shown are from 2 PD-L1lo clones/cell type. Sorted, control CD44+CD133+CD24+ (B16), CD44+CD24+ (ID8agg) and ALDHhi (ES2) TIC formed 50-75% more and bigger spheres vs. PD-L1lo TIC in vitro, consistent with reduced TIC self-renewal in PD-L1lo. In vivo, PD-L1lo B16 and ID8agg TIC exhibited reduced tumorigenicity in wild type mice, consistent with poor TIC function, but immune effects could not be excluded. B16 and ES2 PD-L1lo TIC had significantly reduced tumorigenicity versus control TIC in immune deficient NSG mice consistent with cell-autonomous, immune-independent TIC promotion by PD-L1. In further confirmation, we found reduced serial transplantability of PD-L1lo B16 TIC in NSG mice. The canonical stemness genes nanog and pou5f1 (oct4) were significantly higher in control vs. PD-L1lo TIC for B16 and ID8agg. The canonical stemness gene SOX2 was significantly higher in control vs. PD-L1lo ES2 TIC. Altogether, these data validate effects in human cells, and are consistent with PD-L1 driving TIC through stemness gene control, although specific genes differed in distinct tumors (likely due to differing mutational landscapes). mTORC1 signaling linked to stemness (assessed by raptor) was lower but mTORC2 was unaffected (rictor) in PD-L1lo vs control TIC from all 3 tumors, suggesting mTORC1 control of PD-L1-mediated TIC generation. In support, the mTORC1 inhibitor rapamycin reduced control TIC significantly (>50% vs. untreated). Strikingly, rapamycin significantly increased PD-L1lo TIC in all 3 tumors, suggesting mTORC1 effects differ by tumor PD-L1 status. Interferon (IFN)-γ similarly reduced control TIC and augmented PD-L1lo TIC numbers, suggesting the novel concept that PD-L1 alters IFN-γ signals. PD-L1 sensitization of TIC to IFN-γ and rapamycin is the first example of a molecule enhancing TIC treatment response to our knowledge. We recently reported that αPD-L1 directly reduces B16 and ID8agg proliferation in vitro and in vivo (accession no. 27671674). Preliminary data now show that αPD-L1 (and αPD-1) directly reduce TIC proliferation, and these effects differ from non-TIC. We challenge the paradigm that tumor PD-L1 is primarily an immune escape molecule and that αPD-L1 works primarily to block T cell PD-1 interactions. Citation Format: Harshita B. Gupta, Curtis A. Clark, Bin Yuan, Gangadhara Sareddy, Srilakshmi Pandeswara, Alvaro S. Padron, Vincent Hurez, Jose-Conejo Garcia, Ratna Vadlamudi, Rong Li, Tyler J. Curiel. Cell-intrinsic PD-L1 directs tumor initiating cell fate and rapamycin and interferon-γ response [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5649. doi:10.1158/1538-7445.AM2017-5649

Abstract 1608: CD122-selective IL-2/anti-IL-2 complexes reduce regulatory T cell function and promote CD8+T cell polyfunctionality for durable ovarian cancer immunotherapy

The IL-2 receptor (IL-2R) is an attractive target for cancer immunotherapy as it controls both immune-suppressive regulatory T cells (Tregs) and anti-tumor T cells. We tested depleting Tregs as immunotherapy using anti-CD25 (high-affinity IL-2R subunit) antibodies (αCD25) in ID8agg mouse ovarian cancer (OC). αCD25 reduced ascites and Treg numbers but failed to reduce tumor burden, possibly because it depleted newly activated anti-tumor T cells in tumor-draining lymph nodes. Thus, αCD25 could be novel malignant ascites palliation, but has limited stand-alone efficacy. We then tested IL-2/anti-IL-2 complexes (IL-2c) that selectively stimulate medium-affinity (CD122/CD132) IL-2R thought to expand anti-tumor T cells preferentially, but with little Treg effects. In contrast to several single agents we tested that failed to treat ID8agg (e.g., αCD25, αPD-L1, IL-2 fusion toxin denileukin diftitox), IL-2c alone durably reduced ID8agg tumor burden despite lowering the tumor microenvironmental CD8+/Treg ratio. Thus, we hypothesized that IL-2c improved CD8+ function, reduced Treg function, or both. IL-2c increased polyfunctional IFN-γ+TNF-α+ anti-tumor T cells as expected, an effect that persists weeks after drug clearance. IL-2c also increased anti-tumor T cell CD25 expression that increased IL-2 sensitivity and STAT5 phosphorylation, a likely mechanism for increased polyfunctionality. Unexpectedly, IL-2c reduced the Treg functional mediators CD25, TIGIT and granzyme B, and reduced Treg suppressive function. Thus, favorable Treg modifications are a novel IL-2c mechanism of action. Adding αCD25 to IL-2c to deplete Tregs further unexpectedly worsened IL-2c efficacy in ID8agg and reduced effector memory T cells and polyfunctional T cells in the tumor microenvironment, suggesting a previously unappreciated role for CD25 in IL-2c therapy. Similar data were seen in B16 melanoma, suggesting αCD25 reduction of IL-2c efficacy is not tumor or compartment-specific (ID8agg is peritoneal and B16 is subcutaneous). αPD-L1, an ineffective monotherapy in ID8agg, combined with IL-2c to promote complete responses, suggesting potential for potent, novel combinatorial approaches. Our data suggest that antagonizing high affinity IL-2R (such as to deplete Tregs with αCD25) has limited cancer immunotherapy utility without more specific Treg targeting. In contrast, stimulating medium-affinity IL-2R with CD122-selective IL-2c has great translational promise by simultaneously improving beneficial anti-tumor T cells and reducing detrimental Treg function. Citation Format: Justin M. Drerup, Sri Lakshmi Pandeswara, Aijie Liu, Curtis A. Clark, Alvaro S. Padron, Wanjiao Chen, Vincent Hurez, Tyler J. Curiel. CD122-selective IL-2/anti-IL-2 complexes reduce regulatory T cell function and promote CD8+ T cell polyfunctionality for durable ovarian cancer immunotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1608. doi:10.1158/1538-7445.AM2017-1608

Abstract 3696: Tumor-intrinsic PD-L1 alters tumor chemokines, NK cell trafficking and function, and renders distant PD-L1 null tumors responsive to αPD-L1

Programmed death ligand (PD-L) 1 is expressed on many tumors and inhibits anti-tumor T cells through programmed death (PD)-1. Tumor PD-L1 predicts αPD-L1 treatment effects, but mechanism(s) for PD-L1- tumor response to αPD-L1 are unclear. Our studies suggest tumor-intrinsic PD-L1 signals and spatially varied PD-L1 expression may contribute to response deviation. We used PD-L1+ B16 melanoma (ctrl) and made PD-L1KO by CRISPR. αPD-L1 slowed ctrl but not PD-L1KO B16 growth in mice as expected, but PD-L1KO also responded to αPD-L1 if ctrl B16 was on the trans flank. αPD-L1 elicited similar CD3+ T cell infiltration into ctrl vs. PD-L1KO tumors, but without detectable B16-specific T cell increase. CD11b+ cell infiltration was similar in ctrl and PD-L1KO. Strikingly, NKp46+ and NK1.1+ natural killer (NK) cells infiltrated PD-L1KO > ctrl. NK cells increased significantly (~2-fold) after αPD-L1 in PD-L1KO vs. ctrl, along with NK effector functions (e.g., IFN-γ). Tumor PD-L1 altered tumor chemokines (e.g., CXCL12, CCL2) that could explain trafficking. CD4+ and CD8+ T cells chemotaxed to PD-L1KO slightly > ctrl (with slightly > PD-1+ T cells) in transwells, but NK cells migrated ~2-fold more to PD-L1KO vs. ctrl (with increased activation (e.g., PD-1, CD69), also seen in vivo). We observed that tumor β2 microglobulin (B2M) expression was suppressed by tumor PD-L1. Tumor PD-L1 alteration of immune infiltrates by altering chemokines is a novel mechanism for PD-L1 TIL control. These data also indicate that tumor PD-L1 blunts NK cell infiltration in B16 and reduces B2M, which could boost NK cell anti-tumor activity. These changes correlate with rescued αPD-L1 response of PD-L1KO B16. As αPD-L1 did not increase B16-specific T cells in PD-L1low, we hypothesize that rescue of αPD-L1 response is in part NK cell-mediated. Preliminary in vivo NK cell depletion + αPD-L1 experiments implicate a role for NK cells in PD-L1KO tumor response, but more work is needed to identify mechanisms. Effects extend beyond B16 as ctrl (PD-L1+) MB49 bladder cancer cells in trans elicited αPD-L1 response of PD-L1KO MB49 tumors. Mouse breast and ovarian cancers, and human ovarian and bladder cells exhibit similar PD-L1-intrinsic effects. Finally, αPD-L1 directly altered tumor cell chemokine production in vitro, suggesting additional, novel αPD-L1 treatment mechanisms requiring more study. Our models are useful to understand αPD-L1 (and likely αPD-1) responses based on tumor PD-L1 expression and will help define strategies to improve responses in PD-L1low tumors, and possibly poorly responsive PD-L1+ tumors. Citation Format: Curtis A. Clark, Harshita B. Gupta, Alvaro Padron, Deyi Zhang, Vincent Hurez, Mary Jo Turk, Rong Li, Tyler Curiel. Tumor-intrinsic PD-L1 alters tumor chemokines, NK cell trafficking and function, and renders distant PD-L1 null tumors responsive to αPD-L1 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3696. doi:10.1158/1538-7445.AM2017-3696