Anti-cancer immunotherapy is encountering its own checkpoint. Responses are dramatic and long lasting but occur in a subset of tumors and are largely dependent upon the pre-existing immune contexture of individual cancers. Available data suggest that three landscapes best define the cancer microenvironment: immune-active, immune- deserted and immune-excluded. This trichotomy is observable across most solid tumors (although the frequency of each landscape varies depending on tumor tissue of origin) and is associated with cancer prognosis and response to checkpoint inhibitor therapy (CIT). Various gene signatures (e.g. Immunological Constant of Rejection - ICR and Tumor Inflammation Signature -TIS) that delineate these landscapes have been described by different groups. In an effort to explain the mechanisms of cancer immune responsiveness or resistance to CIT, several models have been proposed that are loosely associated with the three landscapes. Here, we propose a strategy to integrate compelling data from various paradigms into a “Theory of Everything”. Founded upon this unified theory, we also propose the creation of a task force led by the Society for Immunotherapy of Cancer (SITC) aimed at systematically addressing salient questions relevant to cancer immune responsiveness and immune evasion. This multidisciplinary effort will encompass aspects of genetics, tumor cell biology, and immunology that are pertinent to the understanding of this multifaceted problem. Keywords: Cancer immunotherapy, Checkpoint inhibitors, Immune resistance Premise and background prevalence of each landscape may differ. Nevertheless, Anti-cancer immunotherapy is encountering its own this trichotomy is observable across most solid tumors checkpoint. Responses are dramatic and long lasting but suggesting that convergent evolutionary adaptations de- occur in a subset of tumors and are largely dependent termine the survival and growth of cancer in the im- upon the pre-existing immune contexture of individual mune competent host leading to predictable patterns cancers . Current research is trying to determine why determined by uniform immunological principles inde- some cancers respond to CIT more than others and the pendent of the biology pertinent to distinct tumor tissue reasons for individuals’ variability within each indication of origin. It is therefore reasonable to postulate that the [2, 3]. mechanisms leading to cancer resistance to checkpoint Several morphological observations based on immune blockade are similar across cancers deriving from differ- histochemical analyses suggest that three immune ent tissues. Functional characterization based on tran- landscapes best define distinct varieties of the cancer scriptional analyses cannot distinguish structural microenvironment: an immune-active, an opposite differences. Thus a reductionist argument could be made immune-deserted and an intermediate immune-ex- that at the functional level cancers can simply be aggre- cluded . Across cancers, and among subtypes, the gated into immune-active or immune-silent clusters. Current work from our group suggests that most immune excluded cancer resemble functionally immune active * Correspondence: firstname.lastname@example.org tumors suggesting that the periphery immune cells inter- Deepti Kannan and Francesco Marincola were employees of AbbVie at the time of the study but are no longer with the company. act with cancer cells (unpublished observation). Immune-Oncology Discovery, AbbVie, Redwood City, CA, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 2 of 10 We will refer to the mechanisms allowing persistence discrepancy, we surveyed human cancers through of cancer in the immune-active cluster as Compensa- readily-available open-access information. tory Immune Resistance (CIRes) based on the assump- Marincola et al.  have previously described a tran- tion that lack of CIRes would prevent tumor survival scriptional signature comprising the concordant activa- against the host’s immune response. Conversely, we refer tion of innate and adaptive immune effector to survival of cancer in the immune-deserted environ- mechanisms that is required for the occurrence of im- ment as Primary Immune Resistance (PIRes). In 2002, mune-mediated tissue-specific destruction. This repre- we proposed that human cancer immune responsiveness sents a conserved mechanism determining destructive to antigen-specific vaccination administered in combin- autoimmunity, clearance of pathogen-bearing cells dur- ation with systemic interleukin-2 is predetermined by a ing acute infection, acute allograft rejection, graft- tumor microenvironment conducive to immune recogni- versus-host disease and rejection of cancer. Thus, the tion  Likewise, recent observations suggest that CIT is signature was termed: the Immunologic Constant of most effective for the treatment of immune active Rejection (ICR) . The ICR signature was derived from tumors, where a tenuous balance between immune- bulk tumor transcriptome data sets, as they offer the effector and immune-suppressive mechanisms deter- most readily-available sample/data type and the easiest mines outcomes [6–8]. to apply in the clinic due to the ease of collection. The To explain CIRes and PIRes, several phenomenologies ICR signature was further trained to be representative of have been described and models proposed that largely the broader signature as previously described  and is outnumber the fewer immune landscapes (Table 1). currently represented by twenty transcripts and four Such discrepancy can be explained in three ways: a) functional categories: CXCR3/CCR5 chemokines (in- some models do not translate broadly across the major- cluding CXCL9, CXCL10, CCL5), Th1 signaling (includ- ity of human cancers, b) there are subtler immune land- ing IFNG, IL12B, TBX21, CD8A, STAT1, IRF1, CD8B), scapes than those discernable by current approaches, or effector (including GNLY, PRF1, GZMA, GZMB, c) some models are redundant and describe different GZMH) and immune regulatory (including CD274, facets of the same pathophysiology. To solve this CTLA4, FOXP3, IDO1, PDCD1) functions. The expres- sion of these twenty representative genes is highly corre- lated with the extended ICR signature that includes Table 1 Principal models related to immune responsiveness approximately five-hundred transcripts and is represen- Immune Landscape References tative of its main functional orientation as previously de- WNT/βCatenin Silent (0.03) [38, 39] scribed [11, 12]. Importantly, the specific cell types in MAPK Hypothesis Silent (0.001)  the tumor microenvironment expressing these genes will ultimately be relevant in elucidating the mechanistic link Immunogenic Cell Death Active (< 0.001) , [20, 21] between the ICR and the immune responsiveness of can- Regulatory T cells Active (< 0.001) [24, 25] cer. It was subsequently observed that the ICR serves IL23-Th17 Axis Active (< 0.001) [26, 41–44] both as a positive predictor of responsiveness to im- Myeloid Suppressor Cells Active (< 0.001)  munotherapy and as a favorable prognostic marker for PI3K-γ Signature Active (< 0.01) [52–55, 63] various tumor types [6, 10, 13, 14]. This observation sug- IDO/NOS Signature Active (< 0.01) [51, 81, 82] gests that these related phenomena represent facets of a spectrum within the continuum of anti-cancer immune SGK1 Signature Ubiquitous [56, 57] surveillance. Such continuity leads to the fair, though Shc1 signature Ubiquitous  unproven, assumption that signatures predictive of pro- Barrier Molecules Ubiquitous [27, 28] longed survival may mark an immune-favorable cancer Mesenchymal Transition Ubiquitous [29, 30, 83] phenotype and serve as surrogate predictors of respon- Cancer-Associated Fibroblasts Ubiquitous [31–35, 84] siveness to anti-cancer immunotherapies [10, 15]. This TAM receptor tyrosine kinases Ubiquitous [47, 58–60, 85] assumption is also corroborated by recent reports sug- gesting that similar gene expression patterns predict re- Hypoxia/Adenosine suppression Ubiquitous [48, 49] sponse to CIT [6–8]. Specifically, Ayers et al.  using TREX1clearence of Cytosolic DNA NA [86, 87] RNA from pre-treatment tumor samples of Checkpoint Cluster Active (< 0.001) [22, 23] pembrolizumab-treated patients and the nCounter plat- oncogene addicted tumors Silent [11, 68] form identified and validated a pan-tumor T-cell–in- Epigenetic Regulation Ubiquitous [12, 88–90] flamed gene signature correlating with clinical benefit. Distinct models have been assigned to either the Silent or the Active Landscape This tumor inflammation signature (TIS) contains accordingtothe resultsofthe survey showninFig. 1. Ubiquitous refers IFN-γ–responsive genes (CD27, STAT1, IDO1, HLA-E, to models that are not significantly associated with either immune landscape NKG7) related to antigen presentation (HLA-DQA1, Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 3 of 10 HLA-DRB1, PSMB10, CMKLR1) chemokine expression compilation of 10 GEO studies included 1728 cases of (CCL5, CXCL9, CXCR6), cytotoxic activity (CD8A), and breast cancer (compiled in ) that were transcription- adaptive immune resistance (TIGIT, LAG3, CD274, ally characterized utilizing a uniform Affymetrix plat- CD276, PDCD1LG2) and as such, is highly correlated to form. Both datasets were classified according to the the ICR signature: composite scores for each signature coordinated expression of ICR transcripts . calculated with ssGSEA software and compared accord- ICR groups were ranked 1–4, according to the level of ing the expression values in the 999 breast cancer sam- expression of the 20 representative ICR genes (Fig. 1). ples from TCGA were highly correlated (r = 0.98). The At the transcriptional level a dichotomy between TIS has been developed into a clinical grade assay run- Immune-active (ICR3–4) and immune-silent (ICR1–2) ning on the nCounter platform currently being evaluated clusters was apparent . Kaplan-Meier applied to the in ongoing pembrolizumab trials (3). We, therefore, de- four ICR classes confirmed that ICR gene expression veloped a strategy to build a navigational map of can- correlates with survival in breast cancer . cer immunity with the primary purpose of assigning Subsequently, we collected transcriptional Signatures distinct immune responsive and resistant models to their of Responsiveness (or Resistance) (sRes) as reported by respective immune-landscapes using the expression of other investigators (Table 1) and assessed them for their twenty transcripts that are representative of the ex- distribution within the four ICR groups (Fig. 1). The tended ICR signature. signatures tested and respective publication from which Using the ICR signature , we queried the prognos- the transcript biomarkers were derived are available in tic accuracy of a transcriptional data set of breast Table 1. We recognize that the current collection of sRes cancers from The Cancer Genome Atlas (TCGA)asa is far from being comprehensive nor reflective of all pro- discovery platform and validated the findings on a posed models of immune resistance and/or responsive- second transcriptional set of breast cancers from the ness. While further work is being entertained to refine Gene Expression Omnibus (GEO) repository at the and update the collection according to novel under- National Center for Biotechnology Information. The standing of cancer immune biology, for the purpose of TCGA set encompasses RNA-seq-based transcriptional this commentary the current version sufficiently high- characterization of 999 breast cancer cases while the lights the process that we are proposing. Fig. 1 Distribution of sRes gene expression according to distinct models (Table 1) within immune landscapes as defined by ICR gene expression. Four immune landscapes were identified ranked according to the level of expression of ICR genes with purple, green, blue and red representing respectively ICR 1, 2, 3 and 4. Because of similarities in patterns of gene expression, for the purpose of discussion the landscapes will be referred to as immune-silent (ICR1–2) or Immune-active (ICR3–4). Genes were assigned to distinct landscapes according to significant difference in expression between ICR4 and ICR1 (p-value < 0.05 and false discovery rate < 0.1). Genes signatures associated with a particular immune responsiveness model as per Table 1 were assigned to distinct landscapes according to gene enrichment analysis and ubiquitous were considered signatures that did not reach significance (one-tailed Fisher Test p-value < 0.01). *Cluster of ubiquitous genes that segregate with the immune active signatures but did not reach significance and, therefore, were considered ubiquitous Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 4 of 10 Self-organizing clustering of sRes signatures demon- In an effort to move these in silico observations strated a preferential distribution of immune suppressor toward clinical validation and novel biology-based strat- activities such as those related to Th17-IL23 axis, T regu- egies of immune-modulation, new molecular tools which latory cells, checkpoint cluster, myeloid suppressor cells, can be reproducibly applied in the clinic are needed. A IDO within the ICR4 and, to a lesser degree, the ICR3 im- possible candidate is the PanCancer IO 360 Gene mune landscapes (Fig. 1). This finding defines an immune Expression Panel (Nanostring), which allows for multi- phenotype of breast cancer enriched in concert with im- plexed targeted exploration of genes involved in the mune effector and immune suppressive mechanisms. Not tumor-immune microenvironment, allowing for a multi- surprisingly, the transcriptional signature representative of faceted characterization of disease biology and interroga- immunogenic cell death was included in the immune ac- tion of mechanisms of immune evasion. This panel was tive landscape. This information presents a strong argu- developed specifically for translational research and ment for the existence of CIRes mechanisms balancing incorporates many of the PIRs and CIRes signatures in- immune pressure in these cancers’ evolutionary processes. cluding the ICR and the TIS. Conversely, the immune depleted landscapes (ICR1 and ICR2) belonging to the immune silent cluster were Discussion best explained by PIRes, lacking evidence for the prim- Several models have been proposed to explain proclivity ing of a genuine immune response. The sRes of this or resistance of cancer in response to immunotherapy cluster is enriched with transcripts in the PI3Kγ/SFK/ (Table 1). Effector T cell exhaustion is broadly observed pGSK3/β-catenin axis, and activation of the signal trans- in the tumor microenvironment manifesting through the ducer and activator of transcription (STAT3). Coinci- expression of a cluster of immune checkpoints often dentally, these sRes are also associated with suppressive concomitantly expressed in response to chronic inter- myeloid cell differentiation and activation of the IL-23/ feron stimulation [22, 23]. In addition, it is well estab- Th17 axis. However, activation of the PI3Kγ/SFK/ lished that regulatory T cells balance immune effector pGSK3/β-catenin axis does not correspond to activation mechanisms [24–26]. Other models propose blockade of of immunologic transcripts within the same cluster. immune cell homing to cancer tissue by barrier mole- In conclusion, this survey suggested that: cules, chemo-inhibitory mechanisms, and by epigenetic silencing of chemokines (CCL5, CXCL9,and CXCL10), 1. In immune active tumors, signatures of immune Th1 signaling molecules and antigen processing machin- suppression and activation are both present and ery components [12, 27–37]. this balance is responsible for CIRes in the ICR4, Other immune resistance models point to alterations of and to a lesser degree the ICR3, subclasses of breast cancer cell signaling that result in secondary dysregulation cancer. of myeloid cell function. Cancer-intrinsic β-catenin signal- 2. Immune active tumors (ICR3–4) are enriched in ing defects disrupt chemo-attraction of dendritic cells sRes and immunogenic signatures enriched for: (DCs) and, consequently, antigen presentation in the con- a. Immunogenic Cell Death activation text of immunogenic cell death [21, 38–40]. In addition, b. IL23/Th17, polarization of DCs toward a tolerogenic, IL23 producing c. Checkpoints cluster phenotype leading to Th17 polarization was described d. Myeloid suppressor cells in experimental animal models and in human samples e. Regulatory T cells [26, 41–46]. Suppression of anti-cancer immunity has f. IDO also been attributed to the TAM receptor tyrosine 3. Immune-silent tumors are enriched with signatures kinase family members that mediate efferocytosis and reflecting activation of STAT3 and the PI3Kγ/SFK/ negative regulation of DC activity . Similarly, hyp- pGSK3/β-catenin axis and their depletion of oxia can drive immune suppression by inducing tol- immune regulatory mechanisms argues for PIRes: erogenic myeloid DC polarization [48, 49]. Finally, a. β-catenin myeloid cell biology is responsible for the immune b. MAPK activation regulation of the cancer microenvironment through the upregulation of metabolizing enzymes such as ar- Thus, the various models of immune resistance ginase and indoleamine 2,3-dioxygenase, which can (Table 1) converge either into PIRes or CIRes. Inter- negatively impact T cell function [50, 51]. estingly, the CIRes signatures are co-expressed with The phenotype of suppressive myeloid cells in the micro- those reflecting STING activation [17, 18]and im- environment is often attributed to activation of the PI3Kγ/ munogenic cell death [19–21]. This observation sug- SFK/pGSK3/β-catenin axis (Fig. 2). Phosphoinositide3- gests that immunogenicity must be balanced by kinase-gamma (PI3Kγ) can act as a molecular switch that immune suppression in immune active tumors. triggers immune suppressive mechanisms in myeloid DCs Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 5 of 10 Fig. 2 Dichotomy in the Myeloid-Centric Hypothesis of immune resistance: the same pathway is relevant to myeloid cell differentiation as well as intrinsic oncogenic activation (in red boxes are included models included in Table 1). It is currently unclear how the two interpretations diverge vs relate to each other and further characterization of the single cell level will need to be entertained to clarify this point [52, 53]. At the same time, alteration of PI3K functional Therefore, it may be that most models of immune components plays a widespread role in tumorigenesis . resistance are based on a diverse interpretation of the Downstream phosphorylation of serum and glucocorticoid disruption of the PI3Kγ/SFK/pGSK3/β-catenin pathway: kinase 1 (SGK1) by the PI3K/PDK1 cascade leads to activa- one centered on tumorigenesis and the other on myeloid tion of glycogen synthase kinase 3 beta (GSK3β)and subse- cell biology however it is currently unclear whether the quently β-catenin [55–57]. Interestingly, most studies two mechanisms are mutually exclusive or can be ob- describing dysregulation of the PI3Kγ/SFK/pGSK3/β-ca- served in association in the immune active tumors. This tenin axis refer to abnormalities intrinsic to tumor cells, question can only be solved by morphological documen- although the same pathway can play an important role in tation of cell-specific activation of the pathway either by myeloid suppressor DC induction and immune suppression immunohistochemistry or by single cell transcriptional downstream of the TAM receptor tyrosine kinases [58–60]. analysis. However, according to our results and the pub- Converging on the same pathway, hypoxia inducible lished literature [10, 38, 67], it appears that the former factors (HIF1α) signal through the SGK3β/β-catenin interpretation pertains most prominently to the immune axis promoting cancer cell stemness and immune silent cluster (PIRes) while the latter appears to be most suppression [48, 49, 61](Fig. 2). likely pertaining to the immune active (CIRes, Fig. 2). An upstream inducer of PI3Kγ stimulation is the scaffold These results may bear remarkable impact for the design protein SHC1 that shifts the balance between STAT1 and of combination therapies. It appears that, at least in breast STAT3 activation in favor of the latter, promoting immune cancer, therapeutic combinations directed against immune suppression . The context in which activation of SHC1 regulatory mechanism (i.e. checkpoint blockade, IL-23/ preferentially regulates myeloid DC polarization versus can- Th17, TAM receptor kinases, hypoxia factors or IDO in- cer cell signaling remains unclear. Similarly, loss of protein hibitors) will modulate and possibly enhance responsive- tyrosine phosphatase non-receptor type 2 (PTPN2)func- ness of cancers with CIRes (immune active cluster) but tion that inhibits PI3Kγ signaling is associated with activa- will be unlikely to work in the context of immune silent tion of the tumorigenic pathway, while at the same time can cancers of the PIRes phenotype unless complimentary ef- modulate T cell function through mDC activation [63, 64] forts are made to disrupt the non-immunogenic landscape and induction of Th17 polarization [65, 66]. Finally activa- to convert it into an immunogenic one. tion of the mitogen-activated protein kinases (MAPKs)pro- We hypothesize that immune silent tumors evolve by grams is consistently observed in immune silent tumors and employing a strictly essential interface of interactions is associated with a respective mutational signature . with the host’s stroma that limits immune cell recognition. Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 6 of 10 This may be due to the selection of a growth process at the crossroad of two biologies by a “Two-Option devoid of immunogenic cell death (Fig. 1). Thus, these Choice”: 1) immunogenic tumors evolve through a “clean” tumors evolve through the selection of cancer cells disorderly trial-and-error accumulation of oncogenic that adopt refined growth mechanisms reduced to the processes generated by their intrinsic genetic instabil- bare necessities of life. Indeed, preclinical and clinical data ity that leads to a broader number of host-immune focused on molecular subtypes of clinically-validated interactions. These tumors can, therefore, only survive oncogene-addicted tumors (e.g., ALK+, EGFR+, in the immune competent host when immune sup- BRAFV600E+, NTRK-rearranged tumors) indicate that pressive mechanisms balance the immune reaction, 2) these tumors often portray minimal CD8+ T cell infiltra- silent tumors follow a more orderly process with a tion along with reduced expression of immunosuppressive sequential accumulation of essential genetic traits and factors [11, 68]. These molecular subtypes of EGFR- can grow undisturbed by the immune system (Fig. 3). mutated or ALK+ non-small cell lung cancer (NSCLC) Since the latter appear to depend on a leaner carcino- serve as a perfect clinically validated example of “clean tu- genesis, it may be reasonable to postulate that disrup- mors” as these tumors usually do not have high muta- tion of this delicate survival skill may induce messier tional burden, occur in younger patients, and in non- cancer biology prone to immunogenic cell death. smokers. This is supported by recent evidence which Whether this is true remains to be tested. Turning an demonstrates that presence of oncogenic driver mutations immune silent into immune active tumor microenvir- in NSCLC, such as EGFR, ALK, ROS1, RET fusions and onment, even temporarily, may serve a critical thera- C-MET exon 14 skipping is associated with lower muta- peutic role opening the door for immunotherapy tional burden (Mohamed E. Salem, ASCO presenta- strategies. This in turn may be critical because suc- tion 2017, http://abstracts.asco.org/199/AbstView_199_ cessful anti-cancer immunotherapy induces durable 184601.html). This hypothesis is further corroborated tumor regression and immune memory more by the observation that these tumors bear a low frequently. prevalence of mutations in oncogenes suggesting a In conclusion, we propose a systematic, hypothesis- more orderly growth process .It is,therefore,rea- driven task force led by SITC to prioritize and address sonable to suppose that the growth of clean (“onco- the salient questions related to cancer immune gene addicted”) tumors is dependent on activation of responsiveness based on a deeper understanding of the specific pathways (e.g. the PI3Kγ/SFK/pGSK3/β-ca- cancer cell biology that orchestrates distinct immune tenin axis) that avoid immune recognition. Thus, we landscapes. The task force should address outstanding propose that the natural history of cancer is shaped questions to identify conserved versus peculiar patterns Fig. 3 The two-option choice or Hobson’s predicament in cancer survival Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 7 of 10 of immune interaction between the host and cancers of Identification of common pathways that could be different ontology. The role that the genetic background interrogated and targeted to better understand and of the host or micro environmental modifiers play in regu- increase immunogenicity among silent or ‘cold’ lating cancer-immune biology should be addressed follow- cancers ing appropriate stepwise approaches . In addition, a Mechanistic understanding of parameters that could deeper understanding of the evolutionary processes shap- predict immune response to different cancer ing the development of cancer in the immune competent immunotherapies host may offer a simplified understanding of conserved Development of animal models that accurately mechanisms of cancer survival and consequently help the reflect the immune landscape in ‘hot’ versus ‘cold’ identification of a broad range of therapeutics that can tar- human tumors get dominant pathways leading to immunogenic cancer cell death. A clearer qualification of the role played by This workshop will be held in tandem with the SITC adaptive versus innate mechanisms in initiating immune Biomarkers Workshop to be held subsequently on May activation should be considered. Two non-exclusive yet 16–17 in the same premise as part of a strong interest divergent lines of thought are raised to explain immuno- by SITC and other organizations  to deepen the un- genic cancer biology: on one side the high prevalence of derstanding of cancer immune biology particularly in as- neo-epitopes predicted by the higher mutational burden sociation with clinical trial development: (SITC observed in immunogenic tumors positions adaptive im- Biomarkers Workshop). mune recognition at the forefront of immune activation [70–74]. Conversely, immunogenic cell death may primar- ily drive inflammation with secondary recruitment of im- Methods mune cells [20, 21, 75, 76]. The role that each mechanism All data download, processing and analyses were done in plays in human cancer biology, and its implication for R programming environment and as described in Hen- therapeutic intervention, remains to be clarified, and bet- drickx et al. . For the unsupervised clustering of the ter integrated tools may improve our holistic understand- TOE genes (Additional file 1), modified distance and ing of the underlying cancer-immune biology thus hierarchical clustering functions were used. Specifically the facilitating novel biology-based combinational therapeutic distance between 2 genes was defined as 1-“Correlation strategies. Coefficient (Spearman)” and for the hierarchical clustering Finally, better in vivo (genetically engineered and/or function “Ward.D2” method was used. syngeneic) rodent models for the screening of thera- Composite correlation between the ICR and the TIS peutic strategies should be better characterized [77–79]. signature was assessed by calculating a cumulative score Some animal models may be reflective of immune- for each gene included in the respective signature using activated landscapes and be most relevant for the defin- ssGSEA method form GSVA package and correlating ition of therapies combining immune modulatory agents. the scores in the breast cancer TCGA data set according Other animal models may more closely resemble the to Spearman Correlation. biology of immune-silent cancers and would be best uti- The metrics used when assigning genes to silent, active lized to identify therapies that can initiate an immune and ubiquitous groups are derived from differential response before immunomodulatory agents are intro- expression statistics between ICR1 and ICR4 samples. The duced sequentially and/or combinatorically. The avail- genes are assigned to the active cluster if they have signifi- ability of complimentary mouse/human paired panels cantly higher expression levels in ICR4 samples (p-value would largely facilitate such efforts. To our knowledge, < 0.05 and FDR < 0.1). Similarly the genes are assigned to little has been done so far to match mouse models to the silent cluster if they have significantly higher expres- corresponding human immuno-oncology phenotypes sion levels in ICR1 samples (p-value < 0.05 and FDR < 0. following the perspective proposed by this unified theory 1). If the genes do not pass these cutoffs they are grouped of everything. as “Ubiquitous”. Geneset enrichment for each signature The Taskforce will define its goals and future activities belonging to individual models of immune resistance in the occasion of a foundational workshop to be held in (Table 1) against ICR1 and ICR4 clusters was assessed San Francisco on May 14–15 2018 (SITC Cancer Im- using one-tailed Fisher’s exact test. mune Responsiveness Workshop). In the analyses and corresponding heatmaps, the genes The topics to be discussed will include: that were identified in multiple signatures were plotted as one, so each gene in the heatmap is unique. When Interactions between tumor evolution in the the ICR direction is inferred for each signature, the immune competent host and the resulting immune repeated genes contributed to each signature with the landscape same statistics. Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 8 of 10 Additional file Received: 22 November 2017 Accepted: 11 May 2018 Additional file 1: List of individual genes and corresponding models related to immune suppression used for the Theory of Everything (TOE). (TXT 11 kb) References 1. Gong J, et al. Development of PD-1 and PD-L1 inhibitors as a form of cancer immunotherapy: a comprehensive review of registration trials and Abbreviations future considerations. J Immunother Cancer. 2018;6(1):8. CIRes: Compensatory Immune Resistance; DC: Dendritic Cell; 2. Spranger S, et al. Tumor and host factors controlling antitumor immunity GEM: Genetically Modified Mouse Models; GEO: Gene Expression Omnibus; and efficacy of Cancer immunotherapy. Adv Immunol. 2016;130:75–93. GSK: Glycogen synthase kinase; ICR: Immunologic Constant of Rejection; 3. Emens LA, et al. Cancer immunotherapy: opportunities and challenges in IDO: Indoleamine 2,3-dioxygenase; MAPK: Mitogen-activated protein kinase; the rapidly evolving clinical landscape. Eur J Cancer. 2017;81:116–29. PI3K: Phosphoinositide3-kinase-gamma; PIRes: Primary Immune Resistance; 4. Chen DS, Mellman I. Elements of cancer immunity and the cancer-immune PTPN2: Protein tyrosine phosphatase non-receptor type 2; SGK: Serum and set point. Nature. 2017;541(7637):321–30. glucocorticoid kinase; sRes: Signatures of Resistance; STAT: Signal transducer 5. Wang E, et al. Prospective molecular profiling of melanoma metastases and activator of transcription 1; TCGA: The Cancer Genome Atlas; TCIA: The suggests classifiers of immune responsiveness. Cancer Res. 2002;62(13): Cancer Immunome Atlas; TOC: Two-Option Choice; TOE: Theory of Everything 3581–6. 6. Ayers M, et al. IFN-gamma-related mRNA profile predicts clinical response to Funding PD-1 blockade. J Clin Invest. 2017;127(8):2930-40. The design, study conduct, and financial support for this research were 7. Prat A, et al. Immune-related gene expression profiling after PD-1 blockade provided by AbbVie. AbbVie participated in the interpretation of data, in non-small cell lung carcinoma, head and neck squamous cell carcinoma, review, and approval of the publication. and melanoma. Cancer Res. 2017;77(13):3540–50. 8. Charoentong P, et al. Pan-cancer Immunogenomic analyses reveal Availability of data and materials genotype-Immunophenotype relationships and predictors of response to The datasets used and/or analysed during the current study are available checkpoint blockade. Cell Rep. 2017;18(1):248–62. from the corresponding author on reasonable request. 9. Wang E, Worschech A, Marincola FM. The immunologic constant of rejection. Trends Immunol. 2008;29(6):256–62. AbbVie disclosures 10. Hendrickx W, et al. Identification of genetic determinants of breast cancer Tolga Turan, Maulik Patel, Mathew Barnes, Sonia Tanlimco, Rongze Lu, Sarah immune phenotypes by integrative genome-scale analysis. Kongpachith, Kyle Halliwill, Douglas Kline, Thomas Hudson, and Josue Oncoimmunology. 2017;6(2):e1253654. Samayoa are all employees of AbbVie. Deepti Kannan and Francesco 11. Akbay EA, et al. Activation of the PD-1 pathway contributes to immune Marincola were employees of AbbVie at the time of the study. Dr. Howard escape in EGFR-driven lung tumors. Cancer Discov. 2013;3(12):1355–63. Kaufman, Dr. Lisa Butterfield, Dr. Alessandra Cesano, Dr. Davide Bedognetti, 12. Guerriero JL, et al. Class IIa HDAC inhibition reduces breast tumours and and Dr. Wouter Hendrickx have no current or prior affiliation with AbbVie metastases through anti-tumour macrophages. Nature. 2017;543(7645): beyond the current study. The design, study conduct, and financial support 428–32. for this research were provided by AbbVie. AbbVie participated in the 13. Galon J, et al. The continuum of cancer immunosurveillance: prognostic, interpretation of data, review, and approval of the publication. predictive, and mechanistic signatures. Immunity. 2013;39(1):11–26. 14. Bedognetti D, et al. Prognostic and predictive immune gene signatures in Authors’ contributions breast cancer. Curr Opin Oncol. 2015;27(6):433–44. Tolga Turan and Deepti Kannan were primarily responsible for the processing 15. Miller LD, et al. Immunogenic subtypes of breast Cancer delineated by gene and analysis of all the data used in this study. Franco Marincola was a major classifiers of immune responsiveness. Cancer Immunol Res. 2016;4(7):600–10. contributor in writing the manuscript. All authors contributed equally to the 16. Nagalla S, et al. Interactions between immunity, proliferation and molecular conception and design, analysis and interpretation of data and have been subtype in breast cancer prognosis. Genome Biol. 2013;14(4):R34. involved in revising it critically for important intellectual content. All authors 17. Barber GN. STING: infection, inflammation and cancer. Nat Rev Immunol. read and approved the final manuscript. 2015;15(12):760–70. 18. Corrales L, et al. The host STING pathway at the interface of cancer and immunity. J Clin Invest. 2016;126(7):2404–11. Ethics approval and consent to participate 19. Galluzzi L, et al. Immunogenic cell death in cancer and infectious disease. The results shown here are based upon data generated by the TCGA Nat Rev Immunol. 2017;17(2):97–111. Research Network: http://cancergenome.nih.gov/. All of the data used for 20. Wegner KW, Saleh D, Degterev A. Complex pathologic roles of RIPK1 and this study was obtained as described in Hendrickx et al., 2017. RIPK3: moving beyond Necroptosis. Trends Pharmacol Sci. 2017;38(3):202–25. 21. Matt S, Hofmann TG. The DNA damage-induced cell death response: a Competing interests roadmap to kill cancer cells. Cell Mol Life Sci. 2016;73(15):2829–50. Dr. Howard L. Kaufman has served on advisory boards for Amgen, Celldex, 22. Benci JL, et al. Tumor Interferon Signaling Regulates a Multigenic Resistance Compass Therapeutics, EMD Serono, Merck, Prometheus and Turnstone Program to Immune Checkpoint Blockade. Cell. 2016;167(6):1540–1554.e12. Biologics. He serves on a speaker’s bureau for Merck but all honoraria are 23. Koyama S, et al. Adaptive resistance to therapeutic PD-1 blockade is returned to Rutgers University. Alessandra Cesano is Chief Medical Officer at associated with upregulation of alternative immune checkpoints. Nat Nanostring whose platform is mentioned in the text. Commun. 2016;7:10501. 24. Abd Al Samid M, et al. Combining FoxP3 and Helios with GARP/LAP Publisher’sNote markers can identify expanded Treg subsets in cancer patients. Oncotarget. Springer Nature remains neutral with regard to jurisdictional claims in 2016;7(12):14083–94. published maps and institutional affiliations. 25. Elkord E, Abd Al Samid M, Chaudhary B. Helios, and not FoxP3, is the marker of activated Tregs expressing GARP/LAP. Oncotarget. 2015;6(24):20026–36. Author details 26. Tosolini M, et al. Clinical impact of different classes of infiltrating T cytotoxic 1 2 Immune-Oncology Discovery, AbbVie, Redwood City, CA, USA. Clinical and helper cells (Th1, th2, treg, th17) in patients with colorectal cancer. Pharmacology and Pharmacometrics, Abbvie, Redwood City, CA, USA. Cancer Res. 2011;71(4):1263–71. 3 4 Immune-Oncology, AbbVie, North Chicago, IL, USA. Tumor Biology, 27. Salerno EP, et al. Human melanomas and ovarian cancers overexpressing Immunology, and Therapy Section, Division of Translational Medicine, mechanical barrier molecule genes lack immune signatures and have Research Branch, Sidra Medical and Research Center, Doha, Qatar. increased patient mortality risk. Oncoimmunology. 2016;5(12):e1240857. 5 6 Nanostring Technologies, Inc., Seattle, WA, USA. UPMC Hillman Cancer 28. Buckanovich RJ, et al. Endothelin B receptor mediates the endothelial Center, University of Pittsburgh, Pittsburgh, PA, USA. Departments of barrier to T cell homing to tumors and disables immune therapy. Nat Med. Surgery and Medicine, Rutgers University, New Jersey, NJ, USA. 2008;14(1):28–36. Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 9 of 10 29. Hugo W, et al. Genomic and Transcriptomic features of response to anti-PD- 59. Zhang B, et al. Mer receptor tyrosine kinase negatively regulates lipoteichoic 1 therapy in metastatic melanoma. Cell. 2016;165(1):35–44. acid-induced inflammatory response via PI3K/Akt and SOCS3. Mol Immunol. 30. Shields BD, et al. Indicators of responsiveness to immune checkpoint 2016;76:98–107. inhibitors. Sci Rep. 2017;7(1):807. 60. Crittenden MR, et al. Expression of NF-kappaB p50 in tumor stroma limits 31. Feig C, et al. Targeting CXCL12 from FAP-expressing carcinoma-associated the control of tumors by radiation therapy. PLoS One. 2012;7(6):e39295. fibroblasts synergizes with anti-PD-L1 immunotherapy in pancreatic cancer. 61. Vadde R, et al. Role of hypoxia-inducible factors (HIF) in the maintenance of Proc Natl Acad Sci U S A. 2013;110(50):20212–7. stemness and malignancy of colorectal cancer. Crit Rev Oncol Hematol. 2017;113:22–7. 32. Kraman M, et al. Suppression of antitumor immunity by stromal cells expressing fibroblast activation protein-alpha. Science. 2010;330(6005):827–30. 62. Ahn R, et al. The Shc1 adaptor simultaneously balances Stat1 and Stat3 33. Ohlund D, Elyada E, Tuveson D. Fibroblast heterogeneity in the cancer activity to promote breast cancer immune suppression. Nat Commun. 2017; wound. J Exp Med. 2014;211(8):1503–23. 8:14638. 34. Ohlund D, et al. Distinct populations of inflammatory fibroblasts and 63. Karlsson E, et al. Loss of protein tyrosine phosphatase, non-receptor type 2 myofibroblasts in pancreatic cancer. J Exp Med. 2017;214(3):579–96. is associated with activation of AKT and tamoxifen resistance in breast 35. Ozdemir BC, et al. Depletion of carcinoma-associated fibroblasts and fibrosis cancer. Breast Cancer Res Treat. 2015;153(1):31–40. induces immunosuppression and accelerates pancreas cancer with reduced 64. Mattila E, et al. Inhibition of receptor tyrosine kinase signalling by small survival. Cancer Cell. 2014;25(6):719–34. molecule agonist of T-cell protein tyrosine phosphatase. BMC Cancer. 36. Vo DD, et al. Enhanced antitumor activity induced by adoptive T-cell 2010;10:7. transfer and adjunctive use of the histone deacetylase inhibitor LAQ824. 65. Wu R, et al. MicroRNA-448 promotes multiple sclerosis development Cancer Res. 2009;69(22):8693–9. through induction of Th17 response through targeting protein tyrosine phosphatase non-receptor type 2 (PTPN2). Biochem Biophys Res Commun. 37. Zheng H, et al. HDAC inhibitors enhance T-cell chemokine expression and 2017;486(3):759–66. augment response to PD-1 immunotherapy in lung adenocarcinoma. Clin 66. Spalinger MR, et al. PTPN2 controls differentiation of CD4(+) T cells and Cancer Res. 2016;22(16):4119–32. limits intestinal inflammation and intestinal dysbiosis. Mucosal Immunol. 38. Spranger S, Bao R, Gajewski TF. Melanoma-intrinsic beta-catenin signalling 2015;8(4):918–29. prevents anti-tumour immunity. Nature. 2015;523(7559):231–5. 39. Corrales L, et al. Innate immune signaling and regulation in cancer 67. Spranger S, Gajewski TF. Tumor-intrinsic oncogene pathways mediating immunotherapy. Cell Res. 2017;27(1):96–108. immune avoidance. Oncoimmunology. 2016;5(3):e1086862. 40. Spranger S, Gajewski TF. A new paradigm for tumor immune escape: beta- 68. Gainor JF, et al. EGFR mutations and ALK rearrangements are associated catenin-driven immune exclusion. J Immunother Cancer. 2015;3:43. with low response rates to PD-1 pathway blockade in non-small cell lung 41. Coccia M, et al. IL-1beta mediates chronic intestinal inflammation by Cancer: a retrospective analysis. Clin Cancer Res. 2016;22(18):4585–93. promoting the accumulation of IL-17A secreting innate lymphoid cells and 69. Wang E, Uccellini L, Marincola FM. A genetic inference on cancer immune CD4(+) Th17 cells. J Exp Med. 2012;209(9):1595–609. responsiveness. Oncoimmunology. 2012;1(4):520–5. 70. Brennick CA, et al. Neoepitopes as cancer immunotherapy targets: key 42. Kirchberger S, et al. Innate lymphoid cells sustain colon cancer through challenges and opportunities. Immunotherapy. 2017;9(4):361–71. production of interleukin-22 in a mouse model. J Exp Med. 2013;210(5): 917–31. 71. Tureci O, et al. Targeting the heterogeneity of Cancer with individualized 43. Alinejad V, et al. The role of IL17B-IL17RB signaling pathway in breast Neoepitope vaccines. Clin Cancer Res. 2016;22(8):1885–96. cancer. Biomed Pharmacother. 2017;88:795–803. 72. Gubin MM, et al. Checkpoint blockade cancer immunotherapy targets 44. Ngiow SF, Teng MW, Smyth MJ. A balance of interleukin-12 and -23 in tumour-specific mutant antigens. Nature. 2014;515(7528):577–81. cancer. Trends Immunol. 2013;34(11):548–55. 73. Ward JP, Gubin MM, Schreiber RD. The role of Neoantigens in naturally 45. Spivey TL, et al. The stable traits of melanoma genetics: an alternate occurring and therapeutically induced immune responses to Cancer. Adv approach to target discovery. BMC Genomics. 2012;13:156. Immunol. 2016;130:25–74. 74. Noguchi T, et al. Temporally distinct PD-L1 expression by tumor and host 46. Dixon KO, et al. Phagocytosis of apoptotic or necrotic cells differentially regulates the transcriptional expression of IL-12 family members in dendritic cells contributes to immune escape. Cancer Immunol Res. 2017;5(2):106–17. 75. Galluzzi L, et al. Activating autophagy to potentiate immunogenic cells. J Leukoc Biol. 2014;96(2):313–24. chemotherapy and radiation therapy. Nat Rev Clin Oncol. 2017;14(4):247–58. 47. Akalu YT, Rothlin CV, Ghosh S. TAM receptor tyrosine kinases as emerging targets of innate immune checkpoint blockade for cancer therapy. Immunol 76. Galluzzi L, et al. Regulated cell death and adaptive stress responses. Cell Mol Rev. 2017;276(1):165–77. Life Sci. 2016;73(11–12):2405–10. 48. Hatfield SM, Sitkovsky M. A2A adenosine receptor antagonists to weaken 77. Mosely SI, et al. Rational selection of syngeneic preclinical tumor models for the hypoxia-HIF-1alpha driven immunosuppression and improve immunotherapeutic drug discovery. Cancer Immunol Res. 2017;5(1):29–41. immunotherapies of cancer. Curr Opin Pharmacol. 2016;29:90–6. 78. Decker WK, et al. Cancer immunotherapy: historical perspective of a clinical 49. Hu CJ, et al. Differential roles of hypoxia-inducible factor 1alpha (HIF-1alpha) revolution and emerging preclinical animal models. Front Immunol. and HIF-2alpha in hypoxic gene regulation. Mol Cell Biol. 2003;23(24):9361–74. 2017;8:829. 50. Munn DH, Bronte V. Immune suppressive mechanisms in the tumor 79. Maykel J, et al. NOD-scidIl2rg (tm1Wjl) and NOD-Rag1 (null) Il2rg (tm1Wjl) : microenvironment. Curr Opin Immunol. 2016;39:1–6. a model for stromal cell-tumor cell interaction for human colon cancer. Dig Dis Sci. 2014;59(6):1169–79. 51. Mondanelli G, et al. The immune regulation in cancer by the amino acid 80. Masucci GV, et al. The need for a network to establish and validate predictive metabolizing enzymes ARG and IDO. Curr Opin Pharmacol. 2017;35:30–9. biomarkers in cancer immunotherapy. J Transl Med. 2017;15(1):223. 52. De Henau O, et al. Overcoming resistance to checkpoint blockade therapy by targeting PI3Kgamma in myeloid cells. Nature. 2016;539(7629):443–7. 81. Munn DH, Mellor AL. IDO in the tumor microenvironment: inflammation, 53. Kaneda MM, et al. PI3Kgamma is a molecular switch that controls immune counter-regulation, and tolerance. Trends Immunol. 2016;37(3):193–207. suppression. Nature. 2016;539(7629):437–42. 82. Liu Q, et al. Melanoma NOS1 expression promotes dysfunctional IFN 54. Daragmeh J, et al. Analysis of PI3K pathway components in human cancers. signaling. J Clin Invest. 2014;124(5):2147–59. Oncol Lett. 2016;11(4):2913–8. 83. Herbst RS, et al. Predictive correlates of response to the anti-PD-L1 antibody 55. Liu P, et al. Targeting the phosphoinositide 3-kinase pathway in cancer. Nat MPDL3280A in cancer patients. Nature. 2014;515(7528):563–7. Rev Drug Discov. 2009;8(8):627–44. 84. Saleh SMI, et al. Identification of interacting stromal axes in triple-negative breast Cancer. Cancer Res. 2017;77(17):4673–83. 56. Di Cristofano A. SGK1: the dark side of PI3K signaling. Curr Top Dev Biol. 2017;123:49–71. 85. Grabiec AM, Hussell T. The role of airway macrophages in apoptotic cell 57. Xiaobo Y, et al. Serum and glucocorticoid kinase 1 promoted the growth clearance following acute and chronic lung inflammation. Semin and migration of non-small cell lung cancer cells. Gene. 2016;576(1 Pt 2): Immunopathol. 2016;38(4):409–23. 339–46. 86. Demaria S, Golden EB, Formenti SC. Role of local radiation therapy in 58. Crittenden MR, et al. Mertk on tumor macrophages is a therapeutic target Cancer immunotherapy. JAMA Oncol. 2015;1(9):1325–32. to prevent tumor recurrence following radiation therapy. Oncotarget. 2016; 87. Vanpouille-Box C, et al. DNA exonuclease Trex1 regulates radiotherapy- 7(48):78653–66. induced tumour immunogenicity. Nat Commun. 2017;8:15618. Turan et al. Journal for ImmunoTherapy of Cancer (2018) 6:50 Page 10 of 10 88. Place AE, Jin Huh S, Polyak K. The microenvironment in breast cancer progression: biology and implications for treatment. Breast Cancer Res. 2011;13(6):227. 89. Adeegbe DO, et al. Synergistic Immunostimulatory effects and therapeutic benefit of combined histone Deacetylase and Bromodomain inhibition in non-small cell lung Cancer. Cancer Discov. 2017;7(8):852-67. 90. Mondino A, Vella G, Icardi L. Targeting the tumor and its associated stroma: one and one can make three in adoptive T cell therapy of solid tumors. Cytokine Growth Factor Rev. 2017;36:57-65
Journal for ImmunoTherapy of Cancer – Springer Journals
Published: Jun 5, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera