`Mechanisms of Resistance to Immune Checkpoint
`Blockade: Why Does Checkpoint Inhibitor
`lmmunotherapy Not Work for All Patients?
`Chartene M. Fares, MD‘; Eliezer M. Van Allen, MDZ; Charles G. Drake, MD, Phi)“; James i’. Allison, PhD‘; and
`Slwen Hu-Lleskovan, MD, Pill?3
`
`marstaso
`
`The emergence of immune checkpoint blockade therapies over the last decade has transformed cancer
`treatment in a wide range of tumor types. Unprecedented and durable clinical responses in difficuEt—to-treat
`cancer histologies have been observed. However, despite these promising long-term responses, the majority of
`patients fail to respond to immune checkpoint blockade, demonstrating primary resistance. Additionaliy,
`many of those who initially respond to treatment eventually experience relapse secondary to acquired re-
`sistance. Both primary and acquired resistance are a result of complex and constantly evolving interactions
`between cancer celis and the immune system. Many mechanisms of resistance have been characterized to
`date, and more continue to be uncovered. By elucidating and targeting mechanisms of resistance, treatments
`can be tailored to improve clinical outcomes. This review will discuss the landscape of immune checkpoint
`blockade response data, different resistance mechanisms, and potential therapeutic strategies to overcome
`resistance.
`
`JNTRUDUGTIUN
`
`lmmunotherapy has recently become a viable option
`for cancer treatment; however,
`the concept of har—
`nessing the immune system to fight malignancy dates
`back over a century.
`In the 18905, Dr. William Coley
`observed improved clinical outcomes in patients with
`cancer who experienced postsurgicat
`infections.
`Based on these observations, Coley experimented by
`introducing bacterial toxins to patients with sarcoma.
`Although resuits were inconsistent, he was able to
`demonstrate tumor regression in a subset of patients.J
`However, with the advent of chemotherapy and ra-
`diothera py, lmmunotherapy went largely overlooked. In
`the 19505, Ehrlich formulated the concept of immu—
`nosurveitlance, which proposed that emergence of
`maiignant cells is a frequent event, but evoiution to
`clinically relevant disease is suppressed by the im—
`mune system uniess immunity is weakened."a Although
`these early hypotheses fueled the field of cancer im-
`munotherapy, better understanding of immune acti~
`vation, regulation, and interaction with tumor cells and
`the microenvironment was needed.
`
`Now we know that the process of T—cell~mediated
`immunity is a complex sequence of events, with
`constant interplay between stimulatory and inhibitory
`signais that promote adaptive responses against for~
`eign antigens while avoiding autoimmunity. Antigen—
`specifr‘c T cells initially undergo clonaiselection, with
`
`subsequent priming and activation following Tsceil
`receptor recognition of corresponding antigens on
`major histocompatibility compiexes (MHCs) expressed
`by antigen—presenting cells. For full activation, a cos—
`timulatory signal is needed between antigen-presenting
`cells and T cells. Afteractivation and proliferation, T coils
`are trafficked to specific sites by foilowing a chemokine
`gradient. Upon encountering cognate antigen on MHCs,
`effector Tcells (Teffs) release interferon gamma (IFN~y)
`and other cytokines, promoting cytotoxicity and tumor
`cell killing. Following cancer ceil eradication, memory
`”f celis form and remain quiescent until antigen re—
`exposure.
`
`Under normal physiologic conditions, immune check
`points function as negative feedback to reguiate in—
`flammatory responses following T—cell activation. The
`CTLA-4 immune checkpoint receptor was first char~
`acterized by Brunet et al3 in the 19803. Seminal work by
`Krummel and Allison“ demonstrated that CTLA—4 on
`T cells competitively binds to B7 ligands on antigen-
`presentlng cells, interfering with CD28 interactions, thus
`preventing costimulation and the priming phase of T—celi
`activation (Fig. 1A). Subsequently, blockade of CTLA-4
`with antibodies demonstrated tumor
`rejection and
`emerged as proof of concept for immune checkpoint
`inhibitors.5 Another immune checkpoint receptor, PD—l,
`was cloned in 19926 with subsequent characterization
`of its ligand, PDut1.7‘g interaction of PD-l with its
`
`
`
`Author affiliations
`and support
`information (if
`applicable] appear
`at the end of this
`artlcie.
`Accepted on May 1'1.
`2019 and published
`at ascnpubscrg on
`May H, 2019:
`Dfll httpsdldulmrgl
`10.1200IEDBKW
`240837
`
`Downioaded from ascopubsorg by 38.142.205.66 on September 23, 2019 from 038.142.205.066
`Copyright © 20 i 9 American Society of Clinical Oncoiogy. All rights reserved.
`
`Genome EX. 1041
`
`2019 ASCO EDUCATIONAL BOOK l asco.org/edbook
`
`147
`
`Page 1 of 18
`
`Genome Ex. 1041
`Page 1 of 18
`
`

`

`Fares at al
`
`PRACTICAL APPLICATIONS
`
`- immune checkpoint inhibitors provide durable
`clinical responses in multiple difficult~to~treat
`tumor types.
`
`- The tumor microenvironment, tumor immuno-
`genicity, antigen presentation, and classic on-
`cologic pathways piay roles in response and
`resistance to immune checkpoint blockade.
`. By understanding resistance mechanisms to
`immune checkpoint blockade, therapies can be
`devetoped to overcome resistance and treatw
`ment failure.
`
`- Combination treatment strategies with immune
`checkpoint inhibitors are being tested in clinical
`trials, with several already in clinical use.
`
`- Response to immunotherapy may be better
`predicted by using a wide set of biomarkers.
`
`inhibits the effector phase of
`ligands, PD—Ll and PD-L2,
`T—cell activation, thus dampening the immune response.10
`Many tumors are now know to hijack this mechanism to
`avoid T—celi killing, and inhibitory antibodies directed against
`the interaction between PD—l and its ligands have demon~
`strated antitumor responses.u
`
`CLINICAL RESPONSE TO EMMUNE CHECKPCINT INHIBITORS
`
`To date, seven immune checkpoint inhibitors have received
`Li.S. Food and Drug Administration approvai: one CTLA—4
`inhibitor (ipilimumab), three Ple inhibitors iniVolumab,
`pembrolizumab, and cemiptimab), and three PD—Ll
`in-
`hibitors (atezolizumab, durvalumab, and avelumah}.
`lpili-
`mumab was the first immune checkpoint inhibitor to gain
`approval in 2011 for the treatment of melanoma.12 In 2014,
`nivolumab and pembrolizumab were approved in mela—
`noma and have now gained indications for use in non—small
`cell lung cancer (NSCLC), renal celi carcinoma, head and
`neck squamous cell carcinoma, urothelial carcinoma, and
`microsatellite instabilityehigh colorectai cancer, among
`several other tumor types.1366 Atezolizumab, avelumab,
`and durvalumab are approved in many of the same his-
`tologies as the PD-l
`inhibitorsama Most recently, cemi-
`piimah was approved for treatment of metastatic cutaneous
`squamous cell carcinoma.43
`
`PDvllPD~L1 checkpoint blockade tend to be proportional to
`their corresponding tumor mutational burden {Th/l8), pre—
`sumably from the immunogenic neoantigens that are rec—
`ognized as foreign by cytotoxic T lymphocytes (CTLs).“‘Ma
`However, tumors with similar TMB can have very different
`response to checkpoint inhibitors, indicating that response
`to immune checkpoint blockade (ICE)
`is complex, het-
`erogeneous, and inconsistent and that additional mecha-
`nisms are at piay. Increased PDle expression has been
`correlated with immune response and is currentiy used as
`a biomarker for iCB therapy in NSCLC and urothelial car—
`cinema”:50 Additionaliy, elevated numbers of
`tumor-
`infiltrating lymphocytes (TlLs) have been noted in re-
`sponsive cancers?"52
`MECHANiSMS CF RESISTANCE
`
`The biggest chailenges for the cancer immunotherapy field
`are to understand the complex resistance mechanisms and
`to develop effective combination strategies to overcome
`resistance. According to the timing of occurrence, re
`sistance can be primary, as in never—responders, or ac-
`quired, which emerges after a period oi
`response.
`Resistance can also be classified as intrinsic or extrinsic to
`tumor cells. intrinsic resistance is seen when cancer cells
`alter processes that are related to immune recognition, cell
`signaling, gene expression, and DNA damage response.
`Extrinsic resistance occurs external to tumor cells throughout
`the T»cell activation process.
`
`Tumor lmmunogenicity
`
`The abitity for tumors to induce adaptive immune responses
`relies on recognition of cancer cells as foreign. High TMB,
`with accompanying elevated neoantigen expression, plays
`an important role in antitumor imrnunity.““-53 With improved
`sequencing techniques, nonsynonymous mutations Were
`found to generate tumor neoantigens that drive cyto-
`toxic responses against cancer ceils.5“-55 Van Alien and
`colleagues52 demonstrated that mutational
`load was sig
`nificantly associated with response to antiMCTLA—4 treat—
`ment in patients with metastatic‘meianoma. Additionally,
`Rizvi et aim-45 showed that response to antimPDal treatment
`correlated with high TMB and neoantigen load in patients
`with NSCLC.
`in keeping with these studies, poorly immu—
`nogenic tumors with low TMB, such as pancreatic and
`prostate cancers, are inherently more resistant to treatment
`with checkpoint inhibition.44
`
`One of the hallmarks of immunotherapy is the dorabiiity of
`the responses that can be translated into survival benefit.
`indeed,
`in approved indications, checkpoint inhibitor im—
`munotherapy protonged survival in patients with responding
`disease, raising the tail of patient survival curves. However,
`only a subset of turner histologies and a small percentage
`of the patients in each histology are responsive to these
`inhibitors. The response rates of different tumor types to
`
`Extrapolating from these data, mechanisms leading to loss
`of neoantigen expression by cancer ceils may result in
`acquired resistance to ICB. The concept of immunoediting
`exemplifies the impact of neoantigen loss on tumor im—
`munogenicity and explains how resistance might be iorrned
`against cancers with high TMB.
`lmmunoeditirig suggests
`that constant interactions betWeen the immune system and
`cancer ceils result in selection of subclones within the tumor
`
`148
`
`2039 ASCG EDUCATEONAL BOOK | ascocrgledbook
`
`Downloaded from ascopubs.org by 38.142.205.66 on September 23, 2019 from 038.142.205.066
`Copyright © 2019 American Society ofCIinicai Oncology. All rights reserved.
`
`Genome Ex. 1041
`
`Page 2 of 18
`
`Genome Ex. 1041
`Page 2 of 18
`
`

`

`Mechanisms of Resistance to immune Checkpoint Blockade
`
`FiGURE 1. T-Gell Activation and
`Cosignating
`
`(A) T—ceil receptor interacts with
`antigen/major histocompatibility
`complex on APCs. Cosiimulaiory
`signal
`is provided by B7/CD28
`interaction for Twcell activation.
`CTLA—4 competes wiih CD28 for
`B? binding, providing uninhibitory
`signais. (B) Costimulatory signals
`currently being targeted to im
`prove T-cell activation.
`(C) Ex"
`pression of colnhlbliory receptors
`leads to Tcell exhaustion. Colnhl—
`bltory receptors serve as thera—
`peutic targets to enhance aniitumor
`immune response.
`Abbreviation: APC,
`presenting cell.
`
`antigen-
`
`
`
`that lack expression of neoantigens, subsequentiy con—
`ferring poor immunogenicity and resistance to lCB.5'5-57
`With increased intratumor heterogeneity, there is greater
`liketihood that a pooriy immunogenic subcione could be
`setected,
`thus decreasing sensitivity to checkpoint in
`hibitionfsafi‘3 A recent study by Anagnostou et al60 showed
`that relapse of NSCLC tumors after treatment with PD-l/
`PD—Ll and CTLA74 inhibitors demonstrated loss of seven
`to 18 putative neoanttgens, supporting the rote of immu—
`noediting in acquired resistance (Tabie 1). Another study
`recentfy showed that expression of tFN—y paradoxically fa-
`ciiitates immunoediting by CTLs, with resulting gene copy
`number alteration contributing to immune resistance.“
`
`Genetic instability due to atterations in DNA repair and
`replication genes can increase immunogenicity through
`high mutational burden with subsequent neoantigen for-
`mation. Patients with melanoma Were found to have better
`response to antifiPD—l
`treatment
`if
`tumor cells Were
`enriched for mutations in BRO/42, an important homologous
`recombination DNA repair gene.” Similar findings were
`demonstrated in ovarian cancer,
`in which BRCAl/Z
`mutated tumors demonstrated high neoantigen ioads.53
`Aiterations in additional DNA damage response genes,
`inctuding ATM, POLE, FANCA, ERCCZ, and MSH6, have
`recently shown corretation with high TMB and improved
`
`clinical outcomes to iCB in urotheiiat cancer.“ Further-
`
`more, tumors with deficiencies in DNA mismatch repair
`genes leading to microsatellite instability demonstrated high
`mutational burden with enhanced response to ICB across
`a wide range of histotogies.55'55
`
`The presence of PD-Ll—expressing cancer celts within tu-
`mors is known to be an important predictor of response to
`ICB therapy and is commonly used as a biomarker.‘37 it has
`been shown thattumors lacking PD—Ll expression generaily
`show inferior clinicai outcomes to :08 compared with those
`with higher levels of ligands"0 However, tumors with absent
`PD~L1 can respond to ICB, as PD-Ll expression‘ can be
`induced upon activation of the lFN response pathway.
`Regardtess of PD~LI expression, tissues that lack TiLs are
`uniikely to respond to ICB. Tumors with larger numbers of
`TlLs demonstrate greater response to 108 and may serve as
`another predictive biomarkerF‘ifl’” A study in patients with
`metastatic melanoma showed that pie-existing tumoral
`CTLs are a quaiification for response to antinPDul therapy.“
`Tumor Microenvironment
`
`The tumor microenvironment (TME) consists of factors ex—
`trinsic to cancer cells, inctuding various immune and stromai
`cells, vascuiature, extraceliutar matrix, and cytokines that
`influence response to therapy. Immuneusuppressive cells,
`
`Domtloaded from ascopubscrg by 38.142.205.66 on September 23, 2019 from 038.142.205.066
`Copyright © 20E9 American Society of Clinical Oncology. All rights reserved.
`
`Genome EX. 1041
`
`2019 A500 EDUCATIONAL BOOK i asco.orgJedbook
`
`149
`
`Page 3 of 18
`
`
`
`
`
`
`
`
`
`
`
`Genome Ex. 1041
`Page 3 of 18
`
`

`

`Fares et al
`
`TABLE 1. Mechanisms of Resistance In PD—Lt—Overexpressed Tumors, Combination JCS, Tumors With Mutated Chromatin Remodeiing Complexes, and
`High TME Cancers
`
`Specific Circumstances Mechanisms of ResistanceWWW
`
`PD-Ll—overexpressed tumors
`Nonreverslble and sevete Tvceli exhaustion
`
`Coexpresston of énhibitory receptors (LAGG, TIM-3, TiGiT, ViSTA, and
`BTLA)122
`
`Decreased ratio of Tits to Tregs and MDSCS
`
`Altered metaboiism through [DO and increased adenostne production
`
`Mutations in PTEN, EGFR, and Mn?“
`Ple and CTLA-4 inhibitor combination therapy
`lmmunoediting with loss of neoantigens‘50
`
`Deletions or mutations in JAKE/2, iFNGRJ/Z, and ii'r‘Ft“a
`Decreased T—cell priming and DC dysfunction
`Aberrant WNT/fl-catenin signaling
`High copy number loss of tumor suppressor genes“1
`Association with neoantigen overexpression by genetic atteratlons in
`Loss~of«function mutations in chromatin remodeler genes (PBRMI, ARIDZ’,
`mammalian SWI/SNF chromatin remodeitng complexes
`and 3RD?) sensitize tumors to ECB and increase accessibility to
`regulatory elements of iFNJyvlnducible genes. Loss of ARIDIA leads to
`increased mlcrosatcltlte instability with Ena bllity to recruit mismatch repair
`genes during DNA repair, increasing mutational burden and neoantigen
`load. Stability of chromatin remodeling complexes in tumozs contributes
`to ICB resistancembm
`
`High mutation overload tumors
`
`Decreased antigen presentation secondaryto MHC, [52—microglobulin, and
`NLRCS alterations“
`
`
`JAM/2 mutations and decreased iFNJy signaling
`Upregulation of alternate inhibitory checkpoints
`
`Abbreviations: TIGIT, T—cell imrnunoreceptotr tyrosine-based inhibition motif domain; VISTA, V—domain immunogtobuiin—containlng suppressor of T-cell
`activation; BTLA, B and T lymphocyte attenuator.
`
`along with inhibitory cytokin'es in the TME, can undermine
`the antitumor
`immune response.7°-” Regulatory T cells
`(Tregs) are known to facilitate self—toterance by suppressing
`Teff function through inhibitory cytokines and direct contact,
`limittng inflammation”?3 Infiltration of tumors by Tregs has
`been observed in many tumor types, suggesting an immu-
`nosuppressive environment in some cancers?“ The ratio of
`Tetfs to Tregs in murine models is associated with response to
`ICB, in that inability to increase Teffs or decrease Tregs may
`result in resistance to immunotherapyF”?
`
`Myeiotdederived suppressor cells (MDSCs) are another type
`of regulatory colt within the TME that can promote immune
`evasion and tumor growth.“-”"9 MDSCs have been shown to
`ptay a role in facilitating tumor invasion, metastasis, and
`angiogenesisfiu'm Clinical studies demonsttate that increased
`presence of MDSCs within the TME correlates with poor
`response to ICB.82 Accordingly, by inhibiting trafficking of
`MDSCs to the TM E, enhanced response to anti—PD-l therapy
`was seen in a murine model of rhabdomyosarcoma.Ba
`
`Tumor-associated macrophages, particularly M2 macro—
`phages, promote tumo;r progression through modifications
`of the TME.“ M2 macrophages are known to stimulate
`tumor cell motitity, angiogenesis, growth, and immune
`
`evasion?5 Consequently, depletion of tumor—associated
`macrophages in several different morine models corte-
`Iated with reduced tumor gronrth.85-57 Moreover, inhibition of
`myeloid growth factor signaling in macrophages circum—
`vented therapeutic resistanca to lCB in a murine model of
`pancreatic cancerF’189
`
`The cytokine milieu within the TME is involved in immune
`cell recruitment, activation, and proliferation, exerting both
`immune stimutatory and suppressive effects.5m Severat
`chemokines, including COLE, CCtl7, CCL22, CXCLB, and
`CXCL12, play a role in recruiting M0803 and “frogs to the
`TME, thus promoting an immunosuppressive climate.83-9‘
`Consequently, inhibition of the chemokine receptor CCR4
`diminished trafficking of Tregs and promoted antitumor
`effectsPZ-E‘3 Alternately, CXCLQ and CXCLlO recruit CTLs to
`the TM E, with subsequent destruction of cancer cells.”95
`Expression of CXCL9 and CXCLlO can be epigeneticaliy
`silenced, reducing "Fth and promoting resistance to ICB.
`Epigenetic modulator therapy in a model for ovarian cancer
`reversed suppression of these chemokines and enhanced
`response to ICE."6
`
`Transforming growth factor beta (TSP—{3) signaling in—
`fluences multiple TME eiements, including colt growth and
`
`150
`
`2019 A800 EDUCATIONAL BOOK i ascocrg/edboc-k
`
`Downloaded from ascopubscrg by 38.142.205.66 on September 23, 2019 from 038.142.205.066
`Copyright © 2019 American Society ofClinical Oncology. All rights reserved.
`
`Genome EX. 1041
`
`Page 4 of 18
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Genome Ex. 1041
`Page 4 of 18
`
`

`

`Mechanisms of Resistance to immune Checkpoint Blockade
`
`differentiation, wound healing, apoptosis, and immuno—
`suppression. TGF—p
`iimits
`immunosuppression through
`inhibition of CTts while upregulating Tregs.” In a murine
`colorectal cancer model, eievaied TGF—p signaling was
`associated with poorly immunogeriic tumors and limited
`response to ICB, indicating resistance.98 in line with these
`findings, improved antitumor response to ICB was seen with
`inhibition of TGF—[il in metastatic urothelial cancer.99
`
`In addition to promoting angiogenesis, VEGF functions as an
`immunosuppressive cytokine and is associated with re
`sistance to ICB. VEGF levels were found to be higher in
`anti—PD-Z therapy nonresponders compared with re-
`spenders.100 In mouse modets, VEGF impeded commitment
`of lymphoid progenitors, reducing progression to the T—celi
`lineage. ‘0‘ Additionally, VEGF signaling reduces trafficking
`and extravasation of CTLs into the TME while it promotes
`infiltration of Tregs through a selective endothelium.102
`Furthermore, VEGF increases expression of inhibitory re“
`ceptors, contributing to Cit exhaustionm Corroborating
`this evidence, inhibition oi VEGF was correlated with im~
`proved response to ICE in renai ceil carcinoma.104
`
`and CTLA—ti blockade through inability to upregulate M1404
`and Pi)»L1 expressionm’“Mm
`
`Overexpression of alternate immune checkpoints has been
`iinked to antl—PD—t and anti—CtLAwdi
`therapeutic failure.
`Adaptive resistance to ICB was observed secondary to com“
`pensatory upregulation of alternative Immune checkpoint re-
`ceptors,
`including Tncell
`immunoglobuiin, mucin domain-3
`protein (TIM-3), and lymphocyte—activation gene 3 (LAGu3i,
`across multiple studiesi‘iz'124 Alternate immune checkpoint
`receptors continue to be discovered (Fig. 10), inciuding B and
`T lymphocyte attenuator
`(BTLA), T~ceil
`immunoreceptor
`tyrosine‘based inhibition motif domain (TlGiT), and V-dornain
`immunoglobulincontaining suppressor of T—cell activation
`{VISTA}.‘2“27 Coexpression of multiple immune checkpoints
`has been associated with a severely exhausted T-cell state.
`Thommen etaliza demonstrated a positive correlation between
`progressive T-ceil exhaustion and increased coexpression of
`Ple, CTLA~4, TIMVB, LAG-3, and BTLA, with subsequent
`resistance to ICB in NSCLC. Thus, these alternative immune
`checkpoint receptors may serve as potential therapeutic taru
`gets for blockade.
`
`Antigen Presentation and Evolution of Immune Responsa
`
`The evolving immune response, from initial antigen expo“
`sure to cancer ceil cytotoxicity and memonr Twceli formation,
`can be manipulated to evade antitumor immunity. The in—
`ability of T cells to proliferate and adequately diversify likely
`contributes to ECB resistance.
`impaired priming of naive
`T cells through suppressed dendritic cell (DC) recruitment
`was associated with tack of TELs and ICB resistance in
`
`melanoma.”106 Deficiencies in antigen presentation have
`been shown to piay a role in ICB resistance. Multiple studies
`demonstrated that downregulation of MHC ciass | (MHGE)
`allows tumor cells to resist immune surveiliance. 101“” Loss of
`iunction of BE-microglobulin results in disruption of Mi—lcel
`folding and transport to the cell surface, thus mediating
`immune evasion of tumor cellsmgd11 An important study of
`patients with melanoma found truncating mutations in [32
`microglobulin, leading to loss of MHC—l expression and ac-
`quired resistance to iCB.”2 Additionally, mutations within
`the T-cell
`receptorr binding domain of MHC~I haVe been
`identified in colorectal cancer, abrogating cytotoxicity and
`contributing to immune escape.“
`
`The lFN—y signaling pathway mediates immune response
`through the iAKISTAT iamiiy of receptors and transducers.
`iFN-qr signaling upregulates expression of MHC~I, resulting in
`enhanced antigen presentation (Fig. 2A).1m However, IFN—y
`aiso functions within a negative—feedback loop to increase
`expression oi PD—Ll, conferring adaptive resistance to tumor
`celled“?!116 In the context of PDul blockade, amplification of
`PD-Ll
`in Hodgkin lymphoma correlated with Improved re-
`sponse to therapy.117 Multiple studies have demonstrated
`that toss of iAK/STAT signaling results in resistance to PD—l
`
`in addition to expression of inhibitory receptors, exhausted
`T cells demonstrate impaired effector function and altered
`transcriptional state compared with Teffs. T—cell exhaustion
`presents as a spectrum, with association seen between pro
`gressive loss of function and antigen persistence.“’-°’ Chronic
`exposure to cognate antigen also results in elevated PD—l
`expression, with subsequently impaired T—cell
`function.130
`Studies have shown that tumors with low or intermediate
`
`expression of PDsl can be reinvigorated with ICB. However,
`high expression of PD—l was correlated with accumuiating
`T—cell exhaustion and poor response to therapym-m Re-
`cently, epigenetic changes were iinked to T—cell exhaustion, in
`that exhausted cells were found to have a unique chromatin
`landscape that influenced transcriptional state and limited
`effector functionml35 Moreover, the type of distinct chromatin
`state determined if exhausted T cells could be reprogrammed
`after therapy to avoid terminal exhaustion.236
`
`Following effector activity, a minority of T cells enter
`a memory phase,
`remaining quiescent until antigen
`rechalierrgedfl138 Chronic antigen exposure renders pre—
`cursor memory T cells exhausted, with eventual deletion
`and lack of memory formationm-i‘m Given that success of
`ICB is highiighted by marked response durability, memory
`T—cell formation plays an Important role in avoiding re-
`currence and resistance following cessation of treatment.
`Accordingly, patients who responded poorly to anti—PDml
`therapy were shown to harbor fewer tumor-associated
`memory T celis compared with responsive patients.”1
`
`Classic Oncologic Pathways
`
`Through aberrations in oncogenes and tumor suppressors,
`oncologic signaling pathways can regulate immune
`
`Downloaded from ascopubs.org by 38.142.205.66 on September 23, 2019 from 038.142.205.066
`Copyright © 2019 American Society of Clinicai Oncology. All rights reserved.
`
`Genome EX. 1041
`
`2019 ASCO EDUCATIONAL BOOK | ascoorg/edbook
`
`151
`
`Page 5 0f 18
`
`Genome Ex. 1041
`Page 5 of 18
`
`

`

`Fares et al
`
`
`
`FIGURE 2. Cell Signaling Pathways
`Influencing immune Response
`
`Interferon gamma signaling
`(A)
`through .iAKfSTA'i upregulates
`both antigen presentation and
`PD-Ll expression. Loss of MK!
`STAT confers resistance to ICB.
`(B) Aberrant signaling through
`canonical oncologtc pathways
`toads to tumor formation and can
`be targeted to oveicorne immune
`checkpoint blockade resistance.
`
`response (Fig. 2B). These canonical pathways influencethe
`TME by altering immune cell composition and cytokine
`profile, rendering tumors resistant to ICB.
`
`The mitogen-activated protein kioase (MAPK) pathway is
`involved in various cellular activities,
`including proa
`literatioo, apoptosis, and motiiity. Thus, abnormaiities in
`this pathway promote oncogenesis in multiple tumor
`types.“2 increased MAPK signaling impairs recruitment
`and function of TlLs through expression of VEGF and
`multiple other inhibitory cytokines,
`iesulting in immune
`evasion.143 Multipie studies in mouse models have demon
`strated that MAPK inhibitors enhance TlLs,
`lFN-ry signaling,
`MHC~| expression, and PDsLI levels, thus promoting tumor
`celi cytoto><icity.1‘”""146
`
`Loss of the tumor suppressor phosphatase and tensin ho—
`molog (PTEN)
`results in constitutive activation of
`the
`phosphatidylinositol Bakinase {Pl3K) pathway,
`leading to
`tumorigenesis.W Additionally, PTENdeietion was shown to
`decrease CTL activity through expression of VEGF and re
`cruitment of inhibitory cells to the TM 5143-149 Consequently,
`toss of PTEN was associated with resistance to anti—PD—l
`
`therapy in uterine leiomyosarcoma.15° in addition, a PI3K-y
`inhibitor was shown to decrease MDSCs in the TME and
`improve response to ICE in animal models.151
`
`WNT/fl-catenin signaling is another canonical oncogenic
`pathway that is involved in many essentiai ceilular pro—
`ceases“? Abnormalities in WNT/[i—catenin signating have
`been observed in muitipie tumor histoiogies and are
`
`152
`
`2019 ASCO EDUCATIONAL BOOK | ascoorgledhcuk
`
`Downioadcd from ascopubscrg by 38.142.205.66 on September 23, 2019 from 038.142.205.066
`Copyright © 2019 American Society of Ciinica! Oncology. All rights reserved.
`
`Genome EX. 1041
`
`Page 6 0f 18
`
`Genome Ex. 1041
`Page 6 of 18
`
`

`

`Mechanisms of Resistance to Immune Checkpoint Blockade
`
`correlated with increased ceil invasiveness and metastatic
`
`potential.153 in addition to oncogenesis, constitutive WNT
`signaling with stabilization of pwcatenin might be involved in
`ICB resistance through tumor T—ceil exclusion. Spranger
`er al105 showed a negative correlation between level of
`p~catenin and TlLs, mediated by decreased expression of
`the cytokine CCL4 and inability to recruit CDIOB’“ DCs
`needed for T~celi priming.
`
`Indoieamine 2,3udioxygenase 1 (IDOI) is a rate—limiting
`enzyme that converts tryptophan to its metabolite kynur—
`enine and has been associated with suppression of Teff
`function and resistance to ICB.154"55 Additionatly, accu—
`muiation of kynurenines and depletion of the essential
`amino acid tryptophan lead to immune suppression through
`Tucell energy and apoptosls.‘57 In one study, lDO—irnockout
`mice showed improved overaii survival
`(08) with ICB
`compared with wild—type mice, highlighting the therapeutic
`potentlai of iDO inhibition.153
`Other Resistance Mechanisms
`
`There is now emerging evidence that chromatin remodeling
`is involved in sensitivity and resistance to ICB (Table 1). SWI/
`SNF complexes are chromatin remodelers and are fre—
`quently mutated in a variety of cancers.159 The mammalian
`anaiogs of the SWI/SNF compiex, BRGl-associated factor
`(BAF), and polybromo—associated BAF (PBAF), are im—
`portant tumor suppressors. Although these complexes are
`very similar, BAF exclusiveiy contains ARlDlA/B subunits,
`whereas ARIDZ, PBRMl, and BRO? subunits are unique to
`PBAF.‘60 Loss‘of-function mutations in BAF and PBAF have
`recently been shown to sensitize tumors to PD-l and CTLA—
`4 blockade.161 One study demonstrated that inactivation of
`PBAF subunits increased chromatin accessibility to tran—
`scription reguiator elements oi IFNJywlnducible genes, with
`subsequent expression of CXCLQ/CXCLIG cytokines and
`recruitment of TILs.152 Another study showed that loss of
`ARIDJA increased microsateliite instability through den
`ficient recruitment of mismatch repair genes, enhancing
`mutationai burden and sensitizing tumors
`to PD—Ll
`blockade.163
`
`Under normal physiologic conditions, tumor inflammation
`causes hypoxia and ischemia, promoting generation of
`adenosine through dephosphorylation of adenosine mono~
`phosphate by CD73. Upon interaction of adenosine with the
`A2A receptor found on lymphocytes, effector function is
`suppressed,
`iimiting additional tissue injury and coliateral
`inflammatory damage.“166 Overexpression of CD73 has
`been associated with poor prognosis in multipie tumor types,
`encouraging metastasis and angiogeriesis.157'168 Moreover,
`high expression of CD73 promotes Taceil exhaustion and is
`associated with resistance to roam-”0
`
`Within recentyears, a connection between gut microbiome
`composition and response to iCB has been uncovered.
`
`Greater bacterial diversity and enrichment of specific spe
`cies have shown association with improved response to ICB
`across multipie studies. Relative abundance of Faecalrl
`bacterium and Rumfnoccccaceae were observed in re-
`
`sponders, whereas nonresponders were more likely to
`harbor Eiacterminals-s.”“i73 Moreover, germwfree mice with
`fecal
`transplants from responders to ICB demonstrated
`improved outcomes with anti—PD—Ll therapy.m Antibiotics
`can modify response to treatment by altering relative
`amounts of certain species, yielding either increased sus—
`ceptibility or increased resistance to lCB.mm This assd
`elation between intestinal microbiota and response to ICB is
`iikely due to cross—reactivity between microbial and tumor
`antigens, enhancing DC activation, antigen presentation,
`and inflammatory cytokine productionmrug
`THERAPEUTIC STRATEGIES T0 OVERGUME RESISTANCE
`
`With recent insights gained into mechanisms of ICB re-
`sistance, combination strategies using multiple treatment
`modalities are emerging (Table 2). The rationale behind this
`multimodal approach is based upon potential synergistic
`effects of targeting different
`immune escape pathways,
`resulting in improved response to iCB and better patient
`outcomes.
`
`Increasing Tumor lmmunogenicity and T—Cell Priming
`
`One of the first strategies used to bypass resistance empioys
`combination anti—CTLA~4 and anti—PD-l treatment and has
`already been approved in melanoma, colorectal cancer, and
`renal celi carcinoma, with many trials ongoing?”182
`Longer-term data in patients with advanced melanoma
`showad that the 08 rate at 3 years with combined nivoiuu
`main/Epilimumab treatment was 58%, with nivolumab
`monotherapy was 52%, and with ipilimumab monotherapy
`was 34%.183 Although the study was not powered to
`compare combination therapy versus nivoiumab atone, OS
`was descriptively superior with dual biockade. The clinicat
`benefit derived from ICB combination therapy is likely due to
`complimentary mechanisms, as anti—CTLA~4 plays a role in
`priming of T cells, whereas antiePD-l is invoived in later
`reactivation of effector response. Additionally, anti~CTLA-4
`has been shown to deplete Tregs from the TME and en"
`hence Cit—mediated antitumor immunity through broader
`antigen recognition.184
`
`Combination chemotherapy with PDul blockade is approved
`in NSCLC, with modest improvements in progression-free
`survival and 08.135155 Chemotherapy likely sensitizes tumors
`to iCB by increasing release of antigens upon cancer cell
`death, leading to increased priming of CTLsW‘ Furthermore,
`chemotherapy has been associated with depletion of MDSCs
`and Tregs.“33 Radiation therapy is thought to function simi-
`larly, with increased antigen presentation secondary to tumor
`celi death, promoting an inflamed TME and presumably
`synergy with “38.1" Additionally, radiation therapy increases
`
`Downloaded from ascopubscrg by 38.142.205.66 on September 23, 2019 f