Immunotherapy for Gastrointestinal Malignancies

Gastrointestinal (GI) malignancies (esophageal, gastric, pancreatic, intraand extra-biliary ductal, hepatocellular, and colorectal cancers) are an important cause of cancer incidence and mortality in the US and globally. GI cancers account for 15.4% and 23.8% of incident cancers and cancerrelated deaths respectively in the US alone. Although earlier diagnosis and treatment advances have improved outcomes for some GI malignancies, the need for improved therapies in all disease phases (adjuvant, neoadjuvant and advanced) is paramount. Utilization of monoclonal antibodies targeting against vascular endothelial growth factor (VEGF) and epidermal growth factor receptor (EGFR) has shown the success in selected colorectal carcinoma patients. More investigations of immunotherapy are on going in the treatment of GI malignances with different mechanisms and methods. In this article, we review data for established and evolving immunotherapy-related treatment options in GI malignancies.


Introduction
Gastrointestinal (GI) malignancies refer to malignant neoplasms of the GI tract and accessory organs of digestion system: esophagus, stomach, liver and biliary system, pancreas, small intestine, colon and rectum, appendix and anus.Overall, GI malignancies account for more incident cases and deaths than any other organ system.However, these cancers are highly disparate: involving tumors of various histological types (e.g.adenocarcino-ma vs. squamous carcinoma and others) and subtypes with vastly different incidences, lifetime risks and outcomes as outlined in Table 1.
The primary intently curative treatment option for most GI malignancies is still surgical resection though combined modality therapy (concurrent chemotherapy and radiotherapy) has equivalent outcomes in anal cancer.Adjuvant and/or neoadjuvant chemotherapy or chemoradiotherapy has been shown to improve overall survival in select populations.However, given the absence of a proven screening modality in malignancies other than colonoscopy in colorectal cancer, most patients with cancers from GI system are diagnosed at an advanced stage.Effective screening modalities for cancers and discovering active chemotherapeutic, biologic agents in advanced disease are both areas of active investigational efforts in GI malignancies.

EGFR Inhibition: Cetuximab (Erbitux ® ) and Panitumumab (Vectibix ® )
Epidermal growth factor receptor (EGFR) mediated signaling plays important roles in colorectal cancer (CRC) initiation and progression making EGFR inhibition an attractive target.There have been extensive studies in CRC regarding the efficacy, appropriate subpopulation and toxicity.However, limited studies have been pursued in esophageal, gastric, pancreatico-biliary and/or hepatocellular carcinomas.
EGFR engages several downstream signaling cascades including PI3K (PI3K/AKT/mTOR) and MAP kinase (RAS/RAF/MEK/ERK) pathways which mediate cell differentiation, proliferation, and survival.RAS is a membrane bound protein which exchanges bound GDP for GTP and has intrinsic GTPase activity which ensures selfinactivation by GTP hydrolysis.RAS couples growth factor receptors to intracellular signaling pathways by activating downstream targets such as RAF, ERK1, ERK2 and PI3K that promote cell proliferation [17].Three human RAS genes have been identified: HRAS, KRAS, and NRAS.Oncogenic KRAS typically contain single amino acid substitutions (most frequently, in codons 12/13/61) that produce KRAS proteins with strongly reduced intrinsic GTPase activity resulting in constitutively activated GTP-bound state.Activating KRAS mutations are instrumental in the growth and proliferation of a wide variety of tumor types including melanoma, lung, CRC, thyroid and pancreatic carcinomas with a prevalence ranking from 11% (melanoma) to 95% (pancreatic adenocarcinoma) [18].Uniquely in CRC, the chronological sequence of mutations during the tumorigenic pro- cess determines the eventual phenotype [19].Cetuximab (a mouse/human chimeric, IgG1) and panitumumab (a full human, IgG2) are anti-EGFR monoclonal antibodies (MoAb) that competitively inhibit ligand-receptor binding and GTP phosphorylating, effectively disrupting downstream signaling.Additionally, given IgG1 isotype, cetuximab may activate complement pathway and mediate antibody-dependent cellular cytotoxicity (ADCC).Early phase studies of EGFR inhibition in CRC yielded positive results and prompted phase III studies [20]- [40].Given KRAS' role in mediating EGFR signaling, it was postulated that gain of function mutations would result in constitutively activated KRAS and consequent loss of sensitivity to EGFR inhibition in colorectal carcinoma.Proof of concept was initially provided by retrospective analysis of the NCIC CTG/AGITG CO17 phase III trial.KRAS mutation status was determined in 68.9% of the original cohort, and was fortuitously well-balanced in both arms.Authors reported KRAS mutant patients did not benefit from cetuximab, while KRAS wild type (WT) patients had significantly improved PFS/OS [20] [21].This observation was buttressed by analyses of the CRYSTAL and OPUS studies [26].Other published data supports a lower rate of response to EGFR inhibition in patients with BRAF/ NRAS/HRAS mutations or activating mutations of PIK3CA pathway [41].The FDA and EMA recommend that EGFR inhibitors be utilized only in KRAS WT patients, and NCCN guideline recommends further that EGFR inhibitorsshould only be considered in mCRC patients with KRAS and NRAS WT [42].
Cetuximab was the first EGFR inhibitor to be approved by the Food and Drug Administration (FDA) and European Medicine Agency (EMA) both as a single agent in relapsed/refractory colorectal carcinomaand together with combination chemotherapy in the 1st line setting on the basis of several randomized phase III studies.Similar results have been observed with panitumumab subsequently.These results are discussed below and depicted in Table 2.
Since agents targeting both EGFR (cetuximab and panitumumab) and vascular endothelial growth factor (VEGF) (bevacizumab, ziv-aflibercept) have gained regulatory approval for KRAS/NRAS WT mCRCpatients, the sequence of application of either agent has been in debate.FIRE-3, a phase III randomizedstudy, compared FOLFIRI/cetuximab to FOLFIRI/bevacizumab in 592 KRAS WT patients as the first line therapy.Overall response rate (ORR) and median progression-free survival (mPFS) were similar in both arms, however, the median overall-all survival (mOS) was significantly prolonged with the arm with cetuximabfirst compared to the arm with bevacizumabfirst (28.7 vs. 25.0 months respectively) despite greater treatment intensity in the bevacizumab arm [31].Whether the improved mOS observed was related to cetuximab itself or post-progression therapy remains unclear as final results have yet to be published.CALGB/SWOG C80405 is a randomized phase III study of standard chemotherapy regimens (FOLFOX or FOLFIRI) in combination with either bevacizumab or cetuximab in KRAS WT patients as the first line therapy that completed accrual in 2012, the results of which may clarify this issue.
Depending on whether patients were initially treated with a regimen with an oxaliplatin-backbone (FOLFOX/ XELOX) or an irinotecan-backbone (FOLFIRI/XELIRI), 2nd line therapy typically involves a switch between backbones.In KRAS/NRAS WT patients who did not received 1st line EGFR inhibition, adding either cetuximab or panitumumab is advised.There is no data to guide decision making between cetuximab and panitumumab though the higher rate of cetuximab-related infusion reactions in certain geographical regions (in tandem with increased rates of atopy) is a practical consideration [43].In KRAS/NRAS WT patients who received 1st line EGFR inhibition (either cetuximab or panitumumab), this is typically not continued at the time of progression as cross-resistance is assumed given the similar mechanisms of action.Minimal data is available to address this issue: two clinical trials of panitumumab use in KRAS WT patients who progressed on cetuximab containing regimens arrived at divergent conclusions [44] [45].
Somatic mutations in KRAS (<5% -10%) and BRAF (2%) are unusual events in esophageal and gastric cancers [46].Non-randomized studies suggested added RFS/OS/RR benefit when cetuximab was added to conventional chemotherapy in advanced esophagogastric cancer.However, this was not borne out in 2 phase III randomized trials of gastric cancer (EXPAND) and esophagogastric cancer (REAL-3) (see Table 2) [34] [40].The role of adjuvant EGFR inhibition in advanced esophagogastric cancer is being evaluated in a NCI non-randomized phase II study (NCT01360086, perioperative cisplatin/5-FU with cetuximab).
Although phase II studies in unselected populations were promising, the S0205 phase III study of cetuximab in advanced pancreatic cancer was negative [35].However, both cetuximab and panitumumabin combination with cytotoxic agents have demonstrated benefit in several phase II studies in advancedcholangiocarcinoma [47]- [50].
Initial phase II studies demonstrated tolerability of cetuximab when added to chemotherapy in HCC [51] [52].However, a recent phase II study reported OS results that would be inferior when compared to sorafenib [53].
No further phase III evaluation of EGFR inhibition is planned in this disease.

VEGF Inhibition: Bevacizumab (Avastin ® ), Ramucirumab (Cyramza ® ) and Ziv-Aflibercept (Zaltrap ® )
The ability of tumors to induce angiogenesis is a central concept in cancer proliferation [54].Our understanding of tumor angiogenesis and its inhibition has evolved considerably: rather than directly inhibit tumor growth, VEGF-inhibition may mainly normalize abnormal tumor vasculature and improve delivery of cytotoxic agents [55].
Bevacizumab is a humanized monoclonal antibody that binds circulating VEGF-A, preventing its engagement with downstream VEGF receptors (VEGFR-1/2/3) with multifarious effects including inhibition of angiogenic signaling.Bevacizumab was the first biologic agent approved by the FDA to treat any malignancy in 2004 and EMA approval followed in 2005.Approval centered on the results of a front-line randomized phase III study in metastatic CRC with the combination of cytotoxic IFL (irinotecan/5-FU/leucovorin) chemotherapy plus bevacizumab vs. placebo, in which bevacizumab conferred an OS benefit of 4.7 months over IFL alone (20.3 vs. 15.6 months) [56].A flurry of trials followed, evaluating bevacizumab's role in 1st and 2nd line settings with alternative chemotherapy combinations and in other diseases (non-small cell lung cancer, renal cell carcinoma, ovarian carcinoma, and glioblastomamultiforme).These results are summarized in Table 2 [56]- [69].
The aggregate data demonstrated that bevacizumab is preferred in KRAS mutant mCRCpatients population since they do not benefit from EGFR inhibition.Mounting evidence supports prolonged duration of anti-VEGF therapy in metastatic CRC.NO16966 data suggested further improvement when bevacizumab was continued till overt progression and ML18147 confirmed that bevacizumab continuation past progression improved PFS and OS [59] [60].Based on the negative results of AVANT and NSABP C-08, bevacizumabdid not show obvious benefit in either overall survival (OS) or disease-free survival(DFS) at the adjuvant setting for resected CRC [62] [63].
The role of bevacizumab in gastroesophageal cancer treatment has not been approved based on the 2 published large randomized phase III studies (AVAGAST, AVATAR).The AVAGAST study compared chemotherapy versus chemotherapy plus bevacizumab as the first line in patients with metastatic gastric cancer.Although, the study did not show obvious survival benefit by adding bevacizumab, patients with high baseline plasma VEGF-A levels, and low baseline expression of neuropilin-1 showed a trend toward improved overall survival [64] [106].The other study, AVATAR, did not show benefit of bevacizumabadding to the combination of capecitabine and cisplatin in Chinese patients with metastatic gastric cancer [65].A randomized phase III study (ST03) is evaluating the role of adjuvant VEGF inhibition in EGJ/proximal gastric adenocarcinomas.
Ramucirumab is a fully human monoclonal antibody (IgG1) asVEGFR2 receptor antagonist that prevents the binding of VEGF to VEGFR2-the interaction thought to mediate the bulk of VEGF downstream effects.Ramucirumabmonotherapy modestly improved OS (5.2 months vs. 3.8 months) in 2nd line advanced EGJ/gastric carcinoma compared to placebo after prior platinum-containing or fluoropyrimidine-containing chemotherapy [67].Ramucirumab has demonstrated benefit in the 2 nd line setting when combined with paclitaxel chemotherapy in patients who progressed on prior 1 st line platinum-and 5-FU-based combinations (RAINBOW) [68].The median overall survival was 9.6 months for the ramucirumab and paclitaxel combination compared to 7.4 months for paclitaxel alone (p = 0.0169) with 19% reduction in the risk of death.Prior front-line studies utilizing bevacizumab (AVAGAST, AVATAR) were negative, and it is possible that the benefit seen with ramucirumab is secondary to the greater interruption of VEGF signaling with ramucirumabcompared to bevacizumab [64] [65].
In contrast to bevacizumab and ramucirumab, ziv-aflibercept (Zaltrap ® ) is a fusion protein that acts as a decoy receptor for VEGF-A, VEGF-B, and placental growth factor (PlGF).In vitro studies demonstrated that zivaflibercept bound VEGF-A with 100-fold greater affinity and more potent blockade of VEGFR-1/VEGFR-2 than bevacizumab.The phase III VELOUR trial evaluated the combination of ziv-aflibercept and FOLFIRI in the second line metastatic CRC after progression on prior oxaliplatin-based therapy and reported modest benefit [69].Data to support the use of ziv-aflibercept in the 1 st line setting is lacking-the AFFIRM trial which reported similar PFS between both arms was a non-comparative study that was not powered to evaluate the addition of ziv-aflibercept to FOLFOX-6 compared to FOLFOX-6; hence the trial is better considered as an evaluation of alternative 1 st line VEGF inhibition together with FOLFOX-6 [71] [72].A phase II trial is evaluating the OPTIMOX strategy (FOLFOX-7) in the 1 st line setting in combination with aflibercept (NCT01802684, VELVET).
In HCC, bevacizumab has shownimproved survival singly and in combination in multiple phase II studies.VEGF inhibition is considered to have a role in the systemic treatment of advanced HCC in addition tosorafenib [73] [74].However, further randomized phase III trials are needed to confirm this result.HER2/neu encodes the ERBB2 protein, a member of the EGFR family of receptor tyrosine kinases-all of which comprise an extracellular ligand binding domain, a transmembrane domain, and an intracellular domain.HER2 ligand binding activates multiple downstream signals including via the MAPK (RAS/RAF/MEK/ERK), PI3K/AKT, STAT, phospholipase C γ and protein kinase C pathways.HER2/neu over expression has been observed in approximately 15% -25% of breast cancers and is associated with poorer responses to therapy and a clinically aggressive course.HER2/neu over expression occurs at a slightly lower frequency (12% -22%) in esophagogastric malignancies compared to breast cancers although the prognostic implication is unclear [75]- [77].Incidence between esophageal and gastric malignancies is similar; though among gastric cancer sub-types, HER2/neu positivity is seen more often with intestinal-type than diffuse-type cancers [78].For the 7% -22% of patients whose tumors overexpress HER2 by FISH or IHC, the phase III TOGA study was unequivocally positive and resulted in regulatory approval for 1st line use (see Table 2) [79].TOGA's chemotherapy arm consisted of cisplatin with 5-FU or capecitabine, and the addition of trastuzumab to regimens.Diminished efficacy to trastuzumab may develop secondary to primary or secondary resistance-with primary resistance rates as high as 66% -88% in HER2-overexpressing metastatic breast cancer.Although several mechanisms of resistance in breast cancer have been proposed, no data is available in EGJ/gastric carcinomas.Pharmacokinetic data from phase I/II breast cancer trials suggests that failure to achieve steady-state levels secondary to rapid clearance may contribute to primary resistance [80].HELOISE (NCT01450696) is an international phase III study (NCT01450696) evaluating standard (8 mg/kg loading, then 6 mg/kg q3 weekly) versus high (8 mg/kg loading, then 10 mg/kg q 3 weekly) in advanced HER2+ EGJ/gastric carcinomas.

Trastuzumabemtansine (T-DM1
) is an antibody-drug conjugate consisting of trastuzumab linked to the antitubulin agent mertansine (DM1).T-DM1 improved survival by 5.8 months compared to lapatinib/capecitabine in trastuzumab resistant metastatic breast cancer [81].In advanced HER2+ EGJ/gastric carcinomas, T-DM1 is being evaluated in combination with physician choice taxane (docetaxel or paclitaxel).The phase II/III study (NCT01641939) utilizes a novel adaptive design that will evaluate two schedules of T-DM1 (2.4 or 3.6 mg/kg q3 weekly) and pick the phase III dose of T-DM1 depending on the tolerability at 12 weeks.Pertuzumab is a humanized monoclonal antibody that inhibits HER2 dimerization, as distinct from trastuzumab.Combined HER2 blockade with trastuzumab/pertuzumab was evaluated in both the neoadjuvant (TRYPHAENA) and metastatic settings (CLEOPATRA) in breast cancer with high pathologic complete response (pCR) rates and significant improvements in OS respectively [82] [83].Combined HER2 blockade in HER2+ EGJ/gastric carcinomas with trastuzumab/pertuzumab is under evaluation in a phase III trial (BO25114, NCT01774786).
HER2 inhibition is not being evaluated in other GI malignancies.

Checkpoint Inhibitors: Cytotoxic T Lymphocyte Antigen-4 (CTLA-4) and Programmed Death-1 (PD-1)
Peptide antigens presented in association with major histocompatibility proteins (MHC) by antigen-presenting cells (APC) to T-cell receptors (TCR) triggers antigen-specific T-cell activation.T-cells have evolved a two-step mechanism that requires a second signal to mediate whether the antigen-TCR interaction results in proliferation, cytokine secretion, and differentiation or development of tolerance and anergy.T-cell co-stimulatory and co-inhibitory signals are thus potent homeostatic mechanisms that maintain a balance between effective immune responses and peripheral tolerance.T-cell CD28 is the primary co-stimulatory signal modulator while CTLA-4 (CD152) is the primary co-inhibitory signal regulator for CD4+ T-helper, CD8+ T-effector and CD25+ Foxp3+ regulatory T cells.The functional outcome of the ligand-APC-T-cell interaction depends on the relative engagement between APC B7-1/B7-2 (CD80/86) and T-cell CD28 versus CTLA-4.
In addition to CTLA-4, several T-cell surface molecules participate in negative and positive regulation of T-cell activation.The exact role of these molecules in T-cell priming, growth and survival and more specifically in T-effector function, T-helper differentiation, and memory T-cell sustenance are reviewed elsewhere [84].While CTLA-4 initiates the negative feedback loop in T-cell activation, PD-1 is part of the effector phase of this loop.PD-L1 is ubiquitously expressed on tumors and the PD-1/PD-L1 interaction downregulates T-effector responses possibly through suppression of PI3K/AKT activation [85].
CTLA-4 blockade and PD-1 inhibition were thus attractive targets to augment anti-tumor T-cell immunity in cancer.Two CTLA-4 inhibitors [ipilimumab (Yervoy, BMS) and tremelimumab (CP-675206)] and several PD-1/PD-L1 inhibitors have been developed [nivolumab (BMS-936558, BMS), lambrolizumab (MK-3475, Merck) and MPDL3280A (Roche/Genentech/Chugai)] and are in various phases of clinical testing.Of these, ipilimumab has been approved for the treatment of metastatic melanoma in both the 1st line and relapsed settings following successful phase III trials against chemotherapy (dacarbazine) and vaccine (GP-100) comparators respectively [86] [87].Despite promising phase II results, a phase III study of tremelimumab in advanced melanoma was negative and further interest has stalled [88].
Initial evaluation of PD-1 and CTLA-4 inhibition in GI malignancies centered on pancreatic carcinoma but was subsequently extended to advanced CRC and HCC [89].Microsatellite stability is an important prognostic marker in CRC: tumors with high-degree microsatellite instability (MSI-H) have a better prognosis than tumors with low-degree instability (MSI-L) or stable microsatellite (MSS) status.This may be related, in part, to greater immunogenicity associated with MSI-H tumors [90] [91].Investigational approaches in advanced MSI-H CRC include PD-1 (NCT01876511) or PD-1/CTLA-4 combination (NCT01928394, CheckMate 142).Early phase studies are evaluating this approach in combination with chemotherapy in advanced pancreatic cancer (NCT-01473940-gemcitabine/ipilimumab and NCT01896869-FOLFIRINOX/GM-CSF vaccine) and singly in advanced HCC (NCT01658878).A summary of the ongoing trials is provided in Table 3.

Vaccines
Cancer vaccines aim to produce persistent anti-tumor immunity that result in prolonged durable responses.Vaccines are classified based on the antigen(s) incorporated-whole cell, protein, peptide, recombinant virus, dendritic cell, and naked DNA.Cancer vaccines have been studied in various settings (adjuvant, neo-adjuvant and metastatic) across a gamut of malignancies.It is beyond the scope of the article to discuss these studies in detail; however the NCI experience with cancer vaccination between 1995 and 2010 was reviewed in 2 separate publi- cations and reported low response rates of 3% -4% with only infrequent CRs in select malignancies [92] [93].
As TAA have been identified esophageal (MAGE-A3/4 and NYESO-1), gastric (Her-2/neu), pancreatic (MUC1 and mesothelin), hepatocellular (AFP, GPC3, NY-ESO-1, SSX-2, MAGE-A and TERT) and colorectal (CEA) malignancies, cancer vaccination is an attractive strategy in GI malignancies.Given the lack of proven benefit in the phase III setting, no cancer vaccine is approved in the adjuvant, neo-adjuvant and/or advanced disease settings so far.The panoply of ongoing vaccine trials in GI malignancies is summarized in Table 4.
Algenpantucel-L is an allogeneic whole pancreatic cell vaccine engineered to express α-galactosyl (αGal) epitopes that elicit complement-mediated lysis and antibody-dependent cell-mediated toxicity.GVAX is a tumor cell vaccine created by harvesting allogenic cancer cells with subsequent transfection of the granulocyte macrophage-colony stimulating factor (GM-CSF) gene.The algenpantucel-L phase II study involved 70 patients with resected pancreatic adenocarcinoma who received algenpantucel-L vaccination in addition to 5-FU/gemcitabine based chemoradiotherapy (investigational arm of RTOG-9704) post-operatively.Allowing for inherent biases in cross-trial comparisons, 12-month DFS and OS were improved with the addition of algenpantucel-L compared to 5-FU/gemcitabine chemoradiotherapy arm in RTOG-9704 [95] [96].Algenpantucel-L is pending evaluation in a phase III study comparing 5-FU/gemcitabine chemoradiotherapy with or without algenpantucel-L immunization in resected high-risk pancreatic carcinoma (NCT01072981).
Early studies of GVAX immunotherapy in renal cell cancer and melanoma produced middling results but the approach gained traction when an early study reported durable long-term responses in patients with advanced non-small cell lung cancer [97].Both ipilimumab and GVAX have previously been evaluated in advanced pancreatic adenocarcinoma with negative results previously [98] [99].Recent work has shed light on the extensive immunosuppressive microenvironment encircling the primary tumor in pancreatic adenocarcinoma, and explains the poor results observed with conventional immunotherapeutic approaches in prior studies [100].Paradoxically, given the limited access to host immune cells, pancreatic cells may actually be more sensitive to immune attack; and suggests that strategies that combine vaccination with immune checkpoint blockade may be more successful than unselected vaccination.Results of a recent phase Ib study combining GVAX vaccination and ipilimumab in advanced pancreatic adenocarcinoma patients in the 2nd line setting were promising [101].This has prompted  further investigation in the adjuvant setting with GVAX alone in combination with FOLFIRINOX, RT and low-dose cyclophosphamide (NCT01595321) and in advanced disease-GVAX/ipilimumab with FOLFIRINOX (NCT01896869), and GVAX with CRS-207 (ECLIPSE, NCT02004262).Vaccination is also an exciting option in HCC especially considering the presence of HCC-specific TAA (AFP, GPC3, NY-ESO-1, SSX-2, MAGE-A and TERT).Although the tolerogenic tumor micro-environment and HCC-specific tumor suppressive mechanisms are considerations underlying successful vaccination, several studies have demonstrated the validity of this approach.A Japanese study of 150 patients randomized following curative resection to either observation or vaccination with autologous lymphocytes activated in vitro by IL-2 and anti-CD3 reported improved RFS and disease specific survival [102].More recently, an autologous pulsed dendritic cell (DC) approach demonstrated clinical benefit in advanced disease [103].Ongoing vaccine study strategies include intra-tumoral DC vaccination (NCT01974661), NY-ESO-1 vaccination in combination with sirolimus (NCT01522820) and adjuvant treatment following hepatectomy with cytokine-induced T-cells (NCT-01749865).

Conclusions
With the advent of efficacious cytotoxic options in GI malignancies, immunotherapeutic approaches were not pursued aggressively.However, the development and subsequent success of VEGF, EGFR and HER2 inhibition in several GI malignancies have rekindled interest in MoAbs targeting these axes in several settings.
EGFR inhibition has no approved role in advanced HCC, esophagogastric, and pancreatic malignancies.Phase II studies of cetuximab-chemotherapy combinations in advanced cholangiocarcinoma appear promising but randomized phase III data are lacking.Aside from CRC, VEGF inhibition may have use in HCC though it appears unhelpful in pancreatic malignancies.The data in esophagogastric cancers are mixed.Although bevacizumab did not show the improvement in outcomes, ramucirumab demonstrated the survival benefit in a heavily pre-treated patient population with advanced gastric cancer (REGARD) [67], which suggests that alternative (and possibly more intense) VEGF pathway inhibition can improve outcomes, especially when combined with chemotherapy [68].HER2 inhibition has a defined role in HER2+ esophagogastric malignancies.Continued HER2 inhibition post-progression and combined HER2 blockade are areas of ongoing interest.
MSI-H CRC represents a distinct CRC subtype characterized by TAA expression that elicit a strong local (CD8+ T-cell infiltrates and peritumoral lymphoid nodules) host immune response [104] [105].Immune checkpoints (PD-1 and CTLA-4) inhibitors may circumvent immune evasion that contributes to distant spread and the results of NCT01876511 (PD-1 in MSI-H CRC) and CheckMate 142/NCT01928394 (PD-1/CTLA-4 combination in MSI-H CRC) are eagerly awaited.These observations also provide a strong mechanistic rationale for using a peri-operative neoadjuvant approach in stage II/III MSI-H CRC.
Although many GI malignancies have defined TAA, prior vaccine trials were largely negative.Current vaccine studies are utilizing novel delivery systems [CEA expressing virus-like replicon particles (VRP)]; immunomodulatory strategies (low-dose metronomic cyclophosphamide); and are focusing on more immunogenic subtypes of GI malignancies (MSI-H CRC).
Recent trials have validated diverse immunotherapeutic approaches in GI malignancies beyond MoAbs inhibiting EGFR/VEGF.Existing MoAbs are being exploited in alternative settings and novel vaccine strategies are being developed.Immunotherapy is poised at the forefront of adjuvant, neoadjuvant and advanced approaches for the treatment and eradication of GI malignancies.

Table 1 .
Incidence and survival in GI malignancies.

Table 2 .
Published phase III trials of monoclonal antibodies in GI malignancies.

Table 3 .
Phase I/II trials of co-stimulatory agents in GI malignancies in accrual.

Table 4 .
Trials of cancer vaccines in GI malignancies in accrual.