Combination Strategies to Augment Immune Check Point Inhibitors Efficacy - Implications for Translational Research

Immune checkpoint inhibitor therapy has revolutionized the field of cancer immunotherapy. Even though it has shown a durable response in some solid tumors, several patients do not respond to these agents, irrespective of predictive biomarker (PD-L1, MSI, TMB) status. Multiple preclinical, as well as early-phase clinical studies are ongoing for combining immune checkpoint inhibitors with anti-cancer and/or non-anti-cancer drugs for beneficial therapeutic interactions. In this review, we discuss the mechanistic basis behind the combination of immune checkpoint inhibitors with other drugs currently being studied in early phase clinical studies including conventional chemotherapy drugs, metronomic chemotherapy, thalidomide and its derivatives, epigenetic therapy, targeted therapy, inhibitors of DNA damage repair, other small molecule inhibitors, anti-tumor antibodies hormonal therapy, multiple checkpoint Inhibitors, microbiome therapeutics, oncolytic viruses, radiotherapy, drugs targeting myeloid-derived suppressor cells, drugs targeting Tregs, drugs targeting renin-angiotensin system, drugs targeting the autonomic nervous system, metformin, etc. We also highlight how translational research strategies can help better understand the true therapeutic potential of such combinations.


INTRODUCTION
Immunotherapy is often thought to be a recent discovery. However, if we look at the beginnings of cancer immunotherapy, the first scientific attempts were made Fehleisen and Busch, German physicians who noted significant tumor regression post erysipelas infection. William Bradley Coley, the father of immunotherapy first tried to use the immune system in 1981 for the treatment of bone cancer. His work was largely ignored for more than fifty years when the field of immunology took off with the discovery of T cell existence and their role in immunity. In the 1970s, Donald L. Morton treated melanoma patients with Bacillus Calmette-Guerin (BCG) and 91% of patients treated had disease regression. Then in the 1980s, the first immunotherapy cancer treatment IL-2 was approved for the treatment of melanoma and kidney cancer by FDA (1). The most significant breakthrough in immunotherapy came through James Allison & Tasuku Honjo who discovered CTLA4 and PDL1 respectively as therapeutic targets. That research work has further led to the successful development of new checkpoint inhibitors, Dendritic cell therapy, CAR-T cells, and other adoptive cell therapies, oncolytic viruses which have all brought hope to the future of immunotherapy (2). PD1 and CTLA-4 are the two distinct inhibitory receptors on T cells, which are targeted by monoclonal antibodies for providing a durable antitumor immune response. The efficacy of immune checkpoint inhibitors (ICI) was first noted in metastatic melanoma with anti-tumor immune response and significant improvement in overall survival (OS) of patients received ipilimumab an anti-CTLA4 antibody (3). This led to the accelerated clinical studies using anti-PD1/PDL-1 antibodies and regulatory approval in other malignancies including few hematological malignancies, non-small cell lung cancer (NSCLC), and melanoma (4,5). Currently, six ICIs have been approved by the US FDA, of which five ICIs also received market authorization by the European Medicines Agency (EMA).
Even though these therapies have shown survival benefits with reduced incidence of drug resistance or adverse effects than conventional chemotherapy, not all cancer patients are benefitted from these drugs (6). The ability of the tumor microenvironment (TME) to evade immune responses and thereby generating immunosuppression/immune escape is one of the prime challenges that remain (7). This tumor-derived local environment is known as the immunosuppressive TME which ultimately suppresses the antitumor immune response. Clinical trial data analysis of ICI reveals mainly three kinds of patient response: (a) patients who do not respond initially (innate resistance); (b) those who respond at first but fail subsequently (acquired resistance); and (c) those continuously respond from initially to later stages (8,9). Antigen experienced T-cell activation and proliferation are entailed for the generation of an effective antitumor response. The difference in response is because of the insufficient generation and activity of anti-tumor CD8 T-cells (10). Impaired processing and presentation of neo antigens are other causes for the lack of activation of tumorreactive T cells. Multiple other factors like type of cancer, tumor heterogeneity, multiple lines of treatment, and the immunosuppressive TME generated due to tumor-related and non-tumor related factors result in poor response to ICIs (11). Another important obstacle in reduced ICI efficacy is tumor associated hypoxia due to the structural organization of stroma. This can results in altering the pharmacokinetics of the drug by reducing the target drug disposition. Hypoxia can potentiate tumor invasion, stemness, and metastatic capacity mainly through the activation of hypoxia-inducible factor-1a-mediated hepatocyte growth factor/c-Met pathway (12,13). Immunosuppressive microenvironment-associated macrophages enhance tumor hypoxia and further reduce the efficacy of ICIs (14). These results highlight the necessity of combination therapies that can improve the tumor immunogenicity, changing the immunosuppressive TME and targeting other pathways which potentially inhibit the activation of T cells. There are many drugs including cytotoxic as well as non-anti-cancer drugs which have shown promising results in potentiating the efficacy of ICIs.

MATERIALS AND METHODS
We collected 250 published studies that had used the combination of ICIs and other therapies. From the collected studies duplicates were removed and further assessed for eligibility as shown in Figure 1. The review covered all countries with a web-based search of PubMed/Google Scholar published from 2005 to 2019. A further search was conducted in ClinicalTrials.gov to check for the available clinical trials. A few of the clinical trials were further explored by contacting the study teams or information gathered from recent press releases for the study.

Inclusion/Exclusion Criteria
Original primary research such as experimental, observational, and qualitative studies which are written in English and published in peer-reviewed journals are included in the review. Some of the review articles were also selected if they contained outcomes of previous exploratory studies. Search terms were used to find specific studies related to the subject ( Table 1). Articles that define the mechanism behinds the use of drugs as adjuncts with ICIs were considered eligible.
We mainly focused on the mechanisms and rationale of using an adjunct with ICIs for the combinations already in clinical trials. Adjuncts used only in preclinical studies with no currently available clinical data have been excluded from the review.

Cytotoxic Drugs
Conventional chemotherapy drugs such as anthracyclines, cyclophosphamide, Cisplatin, Oxaliplatin, gemcitabine, temozolomide, and paclitaxel are observed to enhance tumor immunity at lesser doses. These drugs at normal dosage and schedule produce immunogenic cell death (ICD) which results in the tumor antigen release and danger signal generation from dying cancer cells known as damage-associated molecular patterns (DAMPs), which in turn provide tumor-targeting immune responses (15). The immunogenicity of tumor cells is modulated by these drugs via various mechanisms comprising (1) enhanced tumor antigen expression and presentation; (2) down regulating coinhibitory molecules (B7-H1/PDL1) and upregulating costimulatory molecules (B7-1) expressed on the surface of tumor cells which in turn increases the function of effector T-cell; (3) granzyme and perforin-dependent mechanisms increasing T-cell facilitated tumor cell lysis; and (4) decreasing MDSCs and Tregs infiltration in the TME.
In mouse immunization experiments using tumor cells pretreated with chemotherapeutic agents such as mitoxantrone and doxorubicin, effective cancer regression via CRT (calreticulin) expression and HMGB1 secretion was observed (16,17). These findings have been confirmed in clinical studies which show that CRT expression is crucial for better prognosis in cancer patients (18). Many studies over the past decade have shown that different chemotherapy regimens also promote differentiation of Th1/Th2 and amplify the proliferation of Tlymphocytes in solid cancers (renal cell carcinoma, colon cancer, and ovarian cancer) (19,20).
Various clinical trials such as KEYNOTE-189, IMPower-132 which used ICI combination with chemotherapy in nonsquamous NSCLC have shown significant benefits.

Metronomic Therapy
Metronomic therapy, an alternate approach for chemotherapy administration encompasses continuous administration of lesser doses of cytotoxic agents. Various studies have reported significant anti-tumor immune responses, lower therapeutic resistance, and a decrease in tumor vascularization (26)(27)(28). Chemotherapeutic drugs such as methotrexate, cyclophosphamide, etoposide, vinblastine, and paclitaxel are used widely as metronomic chemotherapy in various cancers. Even though high dose chemotherapy target cancer cells and are immunosuppressive, high-frequency low dose chemotherapy is considered to be immunostimulatory by targeting the supporting tumor stroma. Cancer cells in tumors cause the surrounding stroma to release additional angiogenesis stimulators. Angiogenesis in the tumor is a process initiated by growth factors and tumorassociated endothelial cells (TEC). When exposed to cancer chemotherapeutic agents at much lower concentrations than those needed to cause tumor cell damage, TECs lose functionality. This property of low-dose chemotherapy such as vinblastine, doxorubicin, paclitaxel, and carboplatin is documented in many studies (29,30). This can reduce the tumor-resistant clones, by giving combination with other treatment modalities (31).
Chemotherapy administered metronomically reduces the number of immunosuppressive regulatory T cells (Tregs) (32)(33)(34), promote antigen-presenting cell maturations (35), improves the activity of dendritic cells (DCs) (36) and, chiefly enhances the activation and function of cytotoxic CD8+ T cells and NK cells. Moreover, it also targets other TME immunosuppressive components referred to as myeloid-derived suppressor cells (MDSCs) (37)(38)(39)(40). MDSCs are a diverse group of myeloid immune cells characterized by their immature state and ability to suppress T cells. This particular property of metronomic chemotherapy can be used for the modulation of the efficacy of ICIs and many studies have reported positive results. It is proposed that metronomic chemotherapy promotes tumorspecific immune activation; concurrent administration of ICIs could maintain the T cells in an activated state. Combination of low dose cyclophosphamide and oxaliplatin sensitized tumor ICI which are initially refractory by increasing CTLs/Treg ratio in the TME in a murine model of lung adenocarcinoma (41). Similar results were obtained in the murine colorectal cancer model where oxaliplatin augmented the amounts of CTLs and activated dendritic cells (DCs) potentiating the efficacy of anti-PD-L1 therapy (42). Enhanced infiltration of CD8+ and CD4+FoxP3T along with suppression of the CD4+ CD25+ FoxP3+ regulatory T cell function have seen after low dose cyclophosphamide with an anti-PD1 agent and improved tumor-free survival in a model of cervical cancer (43,44). There are clinical studies ongoing, combining a metronomic dose of cyclophosphamide along with ICIs as depicted in Table 2.

Thalidomide and Its Derivatives
Thalidomide was first used for its direct anti-tumor actions on myeloma cells by resulting in cell cycle arrest and anti-angiogenic effect. Thalidomide derivatives such as lenalidomide, pomalidomide were then classified as immunomodulatory agents as they stimulate T cells and natural killer cells (NK) to secrete IL-2 and IFNg and inhibit Tregs. This resulted in the amplification of specific immunity in myeloma (45,46). Lenalidomide in combination with dexamethasone and proteasome inhibitor (Bortezomib and Carfilzomib) is considered as a standard chemotherapy regimen for multiple  myeloma. Some studies show that the immunostimulatory effects of lenalidomide are hampered by the concurrent use of dexamethasone (46). The immunomodulatory effects of these drugs can be exploited to enable an enhanced action with ICIs. An invitro cell line study suggests good results with the combination of ICIs and thalidomide derivatives in antileukemic therapy (47). Despite these effects, FDA on July 2017 placed a clinical hold on three clinical trials of the programmed death 1 (PD-1) inhibitor pembrolizumab in combination with lenalidomide or pomalidomide for patients with multiple myeloma: KEYNOTE-183, KEYNOTE-185, and KEYNOTE-023 because of increasing reports of death. However, there are similar clinical trials ongoing with the combination of ICIs and lenalidomide for other indications such as lymphoma (47)(48)(49). In KEYNOTE-023 responses were durable in the overall study population which includes lenalidomide refractory patients with a median duration of response (DOR) of 18.7 months, Objective response rate (ORR) of 44%. Median PFS was 7.2 months and median OS was not reached (50). The median PFS was 27.6 months with a median follow-up of 32.2 months in a two-year update of a Phase II trial of Pembrolizumab, Lenalidomide, and dexamethasone as post autologous stem cell transplant consolidation in multiple myeloma. Similarly, Badros et al. reported ORR-60%, DOR 14.7 months, median PFS 17.4 months, Median OS not reached with Pembrolizumab, pomalidomide and dexamethasone combination and they also reported long term remissions after discontinuing Pembrolizumab, 57% have ongoing response without any further treatment with a median survival of 18 months (51, 52).

Epigenetic Therapy Hypomethylating Agents
Hypomethylating agents such as azacytidine and decitabine are mainly used in the treatment of myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). Due to the availability of advanced genomic studies, it is evident that hypermethylation of tumor suppressor genes in the promoter region is one of the pathways of carcinogenesis (53). The most common and stable of epigenetic alterations in cancer is methylation of DNA at CpG islands. Hypermethylation tends to limit ICIs by blocking endogenous interferon responses required for cancer cell recognition whereas global hypomethylation marks in the expression of inhibitory cytokines and PD-L1, supplemented by epithelial-mesenchymal variations contribute to immunosuppression. The drivers of these opposing states of methylation are not stated well. DNA methylation is also essential in the 'exhaustion' of cytotoxic T cells that occurs as a result of tumor progression (54). The feasibility of targeting this mechanism with hypomethylating agents is being widely studied now. Other than this, the collective data of studies suggests that hypomethylating agents are known to promote the expression of cancer-specific antigens and major histocompatibility complex (MHC) results in improved immunologic recognition of cancer cells and enhanced antitumor responses (55)(56)(57). This property is exploited By Brahmer J et al. and when six patients in a clinical trial of epigenetic therapy for advanced treatment-resistant NSCLC were switched to an immunotherapy trial targeting the PD-1/PD-L1 immune tolerance checkpoint, they saw a substantial improvement. Three of the six patients had long-term partial responses to immunotherapy, ranging from 14 to 26 months, while the other two had stable disease for 8.25 and 8.5 months, respectively (4,5,58). Various clinical studies are ongoing to check the efficacy and toxicity of this combination therapy in various cancers ( Table 2).

Histone Deacetylase Inhibitors (HDACis)
As discussed above, in cancers that are immune to immunotherapy, epigenetic inhibitors may be used as adjunctive therapy or as a replacement for immunotherapy (sole agents of immune modification). In the TME, epigenetic changes are normal, resulting in a variety of gene expression changes and tumor escape. Recently, HDAC inhibitors have shown potential anti-cancer properties such as cell cycle arrest, anti-angiogenesis, activation of both intrinsic and extrinsic apoptotic pathways, autophagy, and modulation of immune responses. HDACs are classified mainly into 4 categories (Class I, Class II, Class III, Class IV) based on their structure, which is further subdivided into Zinc dependent (Class I, Class II, Class IV) and NAD+ dependent (Class III) for their enzymatic activity (59). Modulation of the immune response can be utilized for improving the outcome of ICIs. HDACis can change the expression of immune system upregulaters including MHC and costimulatory molecules, which affects antigen presentation and thus T cell activation. It has also been observed that naïve T cell functionality is stimulated by HDAC6i and Tregs are targeted by class II HDACis. The Class I HDACis target adaptive immunity by enhancing natural killer and CD8 cells functionality. Several non-selective HDACis such as vorinostat, belinostat, trichostatin A are being tried in clinical trials (60). HDAC I, II, and IV are inhibited by Pan-HDACi such as valproic acid, vorinostat, and panobinostat. MS275 (a class I HDAC inhibitor), MC1568 (a class III HDACi), sirtinol (a class II HDACi), and Nexturastat A (HDAC6i), among many others, are currently being used to reduce the toxic effects associated with pan-HDAC inhibition. The wide range of biological effects may be due to each inhibitor's particular chemical structure and mechanistic profiles (61).
In a B16F10 syngeneic murine model study by Wood DM et al. HDAC inhibitor treatment resulted in rapid upregulation of histone acetylation of the PD-L1 gene leading to improved and persistent gene expression. The effectiveness of combining HDAC inhibitor with PD-1 blockade for the treatment of melanoma was studied in a murine B16F10 model and compared to control and single-agent therapies, they had a slower tumor development and a higher survival rates (62).
Targeted Therapy in Combination With ICIs CDK4/6 Inhibitors CDK inhibitors are a class of naturally occurring molecules that belong to the Cyclin-dependent kinase inhibitors family INK4 and specifically inhibit the CDK4/6 proteins. By dephosphorylating the retinoblastoma tumor suppressor protein, CDK4/6 inhibitors block the G1 to S phase transition of the cell cycle, resulting in cell cycle arrest in tumor cells. Teo ZL et al. showed that a combination of CDK4/6 and PD1 blockade resulted in intensifying tumor growth inhibition. Increased antigen presentation by tumor cells, stimulation of effector T lymphocyte activation, and decreased proliferation of immunosuppressive Treg cells are among the various mechanisms suggested (64,65). The immune checkpoint pathway is majorly implicated with the increased immune response upon inhibition of CDK4/6 inhibition. The results of a phase Ib clinical trial combining abemaciclib and pembrolizumab in ER-positive HER2-negative women at a 16-week interim study, MBC demonstrated that this combination is secure, with an ORR of 14.3% (66).
In the JPCE trial, 25 hormone receptor-positive Her2 negative metastatic breast cancer patients were treated with a combination of Abemaciclib and Pembrolizumab. In this trial, the partial response (PR) rate at 16 weeks was 14.3% compared to 6.8% seen in the MONARCH-1 trial which used single-agent abemaciclib. There were no complete responses in either trial. In both experiments, the incidence of stable disease was 60.7% in JPCE and 60.6% in MONARCH-1. In the JPCE experiment, fewer patients developed a progressive disease (17.9% and 25.0%). At the early assessment, the disease control rate (DCR) in JPCE was 75%, while in MONARCH-1 it was 67.4%. As of 17-04-2020, the study is still ongoing and results are yet to publish (66).

Anti-Angiogenic Drugs
Anti-angiogenic drugs work by inhibiting platelet-derived growth factor receptors (PDGFRs) and fibroblast growth factor receptors, which target vascular endothelial growth factor (VEGF)/VEGF2 or other small molecules involved in angiogenic and proliferative pathways (FGFRs). Preclinical studies have shown that abnormal tumor vasculature promotes immunosuppressive TME, which can be reversed with antiangiogenic therapies (67,68). Inflammatory mediators such as cytokines and immune cells promote and control angiogenesis, which in turn may affect the microenvironment ( Figure 2) (69). Anti-angiogenic agents can thus activate the immune system, and immunotherapy can also have an anti-angiogenic effect, therefore they can act synergistically on the tumor.
Results of the phase 3 IMpower150 (NCT02366143) study reported in 2018 showed significant improvement in OS of patients treated with a combination of Atezolizumab, Bevacizumab along with chemotherapy in treatment-naïve metastatic non-squamous non-small-cell lung cancer patients. A total of 800 patients were given either atezolizumab plus bevacizumab plus carboplatin plus paclitaxel therapy (ABCP group) or bevacizumab plus carboplatin plus paclitaxel therapy (BCCP group) (BCP group). The study found that the ABCP community had slightly longer progression-free survival (PFS) and overall survival (OS) (

Inhibitors of DNA Damage Repair
Synthetic lethal impact on cancer cells with a dearth in homologous recombination (HR) is induced by Poly ADPribose polymerase (PARP) inhibition. In addition, PARP inhibitors (PARPi) can enhance Th1-skewing immunity and modulate the immune microenvironment, as well as promote the priming of anti-cancer immune responses (72). As the immunological effect of PARPi is multifaceted, it might favor an increased efficacy of ICI treatment and boost the cancerimmunity cycle. Olaparib is a first-in-class PARP inhibitor for the treatment of adult patients with germline BRCA mutated advanced ovarian cancer, breast cancer, fallopian tube cancer, and peritoneal cancer. As discussed the immunological effects of PARP inhibitors are multifaceted (73), which includes

Other Small Molecule Inhibitors
Several multiple tyrosine kinase inhibitors have shown enhanced action with ICIs. Sorafenib, regorafenib, and Lenvatinib are some of the drugs which have demonstrated superior effect with ICIs in hepatocellular carcinoma. Hepatocellular carcinomas generally arise from an immunosuppressive environment (77). Since it must interact with a variety of foreign antigens, the liver suppresses immune responses to toxins and antigens draining from the enteric circulation. Because of the progressive nature of the disease and the tolerogenic characteristics of the liver, the HCC TME has immunosuppressive characteristics (77)(78)(79). HCC exploits this immune tolerance to initiate and promote HCC carcinogenesis and progression. Even though the TME is highly immunosuppressive there are reports which show that ICIs have benefits in the treatment of HCC (78)(79)(80). Most of these successful attempts of ICI usage in HCC were demonstrated in combination with multiple tyrosine kinase inhibitors such as sorafenib and regorafenib or when pretreated with these small molecule inhibitors (78)(79)(80). However, the exact mechanisms of their additive activity are yet to be proven.
Other small molecule inhibitors include drugs which target PI3K and MAPK pathway. Both the PI3K-Akt-mTOR and the RAS/RAF/MEK/MAPK pathways have been implicated in the control of PD-L1 expression in previous studies (81,82). Vemurafenib, a BRAF inhibitor, can enhance T cell antigen expression in melanoma and thus stimulate T cell immune responses. By the expression of anti-apoptotic proteins, the PI3K-Akt-mTOR pathway is also involved in resistance to T cell-mediated killing. A more recent study reported better tumor immune infiltration and more anti-tumor immune control in a CD8 T-cell-dependent way when the dual blockade of BRAF and MEK combined with PD-1 inhibitors (83). The experimental evidence presented above may be used to justify clinical trials of BRAF and/or MEK inhibition in conjunction with ICIs. Many studies are ongoing to test these hypotheses and some of the study results are promising. In a study, the selective targeting of PI3K-g with a small molecule inhibitor IPI-549 (NCT02637531) showed that it could restore the TME and overcome resistance to ICIs by inducing cytotoxic T cell-mediated tumor regression (84). This discovery paves the way for the appealing technique of combining PI3Kis and ICIs to prevent immune checkpoint blockade resistance. One or two agents targeting the PI3KAkt-mTOR pathway could be a possible combination partner for immune checkpoint blockade (85), but the most effective combination is not defined.
Cabozantinib a multikinase inhibitor whose targets include MET, AXL, and VEGFR2 in combination with Nivolumab showed better response rate, delayed disease progression, and extended survival in the CheckMate -9ER trial (86). CheckMate -9ER (NCT03141177) is an open-label, randomized, multi-national Phase 3 trial that investigates Cabozantinib+Nivolumab combination in treatment naïve advanced or metastatic renal cell carcinoma. However, 3 drug combinations of Cabozantinib+Nivolumab+Ipilimumab for the same indication were discontinued.

Bruton's Tyrosine Kinase Inhibitors
Bruton's tyrosine kinase (BTK), bone marrow-expressed kinase (BMX), redundant resting lymphocyte kinase (RLK), and IL-2 inducible T-Cell kinase are all members of the Tec kinase family (ITK). BTK is primarily expressed on B cells, though not exclusively, and ITK is primarily expressed on T cells (88).
Ibrutinib is a small molecule that not only targets BTK, which is required for malignant B cell survival, but also blocks ITK, which shifts the Th1/Th2 balance and improves antitumor immune activity. The combination of ibrutinib anti-tumor activity and ICIs showed synergism in preclinical models and its positive therapeutic outcome does not result from the direct action against tumor cells, but rather from their activity on the immune system (89). Ibrutinib activates CD8+ T cells, lowers MDSC cytokine secretion, and modifies the inhibitory activity of immune checkpoint molecules like the PD1/PD-L1 and CTLA-4 axis on tumor-infiltrating lymphocytes (89,90). Many clinical trials are testing this hypothesis and most of them are being tried in lymphoma patients. However, there is limited data in solid tumors and one study showed no benefits of adding ibrutinib along with ICIs (durvalumab) in pretreated solid tumors such as pancreatic cancer (ORR-2%, Median OS-4.2 months), breast cancer(ORR-3%, Median OS-4.2 months) and NSCLC (ORR-0%, Median OS-7.9 months) (91).

Selective Estrogen Down Regulators
Fulvestrant is a selective ER down regulator (SERDs), a class of ERa antagonists. Fulvestrant treatment aids in overcoming various resistance mechanisms by ERa down-regulation. Several preclinical studies have proposed various mechanisms by which SERDs exert an anti-tumor response. It alters the TME by increasing Th1 immune response, decreasing Tregs, and inhibiting estrogen-mediated cell proliferation pathways (92,93). Results of recent translational research indicate that SERDs with strong anti-estrogen activity such as Fulvestrant and potentially other anti-estrogens (94) can augment the action of ICIs to inhibit breast cancer progression. As a result, combining anti-estrogens with ICI may be an effective treatment technique for both ER-positive and potentially ER-negative or treatment-resistant breast cancers, greatly increasing the use and life-extending effects of these drugs to improve patient survival (95).

Combination of Multiple Checkpoint Inhibitors
Recently, many studies are focusing on the combination of immunotherapy agents which target multiple pathways in cancer. It is proposed that targeting multiple checkpoints can increase the activity of each other and thereby overcoming each monotherapy's limitations. The combination of CTLA-4 and PD 1/PD-L1 blockade has exhibited antitumor efficacy in preclinical models (96). The logic of combining multiple checkpoint inhibitors is that they have different mechanisms of action, with anti -CTLA-4 mainly acting in the lymph node compartment which is responsible for restoring the induction and proliferation of activated T cells, and with anti -PD-1 mainly acting at the periphery of tumor site, preventing the neutralization of cytotoxic T cells by PD-L1 expressing tumor and plasmacytoid dendritic cells in the TME (97).
In the RCT checkmate 067 trial in advanced melanoma, patients with PD-L1-positive tumors had the same median progression-free survival (14 months) when they were given nivolumab alone or in combination. However, in patients with PD-L1-negative tumors, ipilimumab plus nivolumab had a longer progression-free survival (11.2 months) than nivolumab alone (5.3 months). The major disadvantage of combination immunotherapy is the increase in grade 3 or grade 4 toxicities. The updated data in 2017 reported that only patients with no expression of PD-L1 benefited from the combination and the recent update in 2019 added that the %age of patients experiencing a complete response continued to increase, with complete response rates at five years of 22% for combination therapy and the proportion of patients alive and treatment-free was 74% in nivolumab/ipilimumab combination (98).
Based on ORR and DOR in phase I/II CheckMate-040 trials, FDA has approved Nivolumab-Ipilimumab combination in Hepatocellular carcinoma in March-2020. In the study, 16 of 49 patients (33%) treated with nivolumab in combination with ipilimumab responded to treatment after a minimum of 28 months of follow-up (95% CI, 20-48). Four patients (8%) had a full response (CR) and 12 people (24%) had a partial response (PR). DORs ranged from 4.6-30.5 months, with 88% lasting at least 6 months, 56% lasting at least 12 months, and 31% lasting at least 24 months (99,100).
Updated result of Phase 3 Checkmate-214 trial (Ipilimumab +Nicolumab) in treatment naïve renal cell carcinoma after 42month median follow-up showed very promising result with CR in 10% patients, OS rate and ORR as 52% and 42% respectively and DOR was not reached (101).
FDA granted fast track designation for investigation of the combination of Balstilimab PD-1 inhibitor and Zalifrelimab a CTLA-4 inhibitor for the treatment of patients with relapsed or refractory metastatic cervical cancer. The FDA designates an investigational drug for fast track review in order to speed up the production of drugs that treat severe or life-threatening conditions and address an unmet medical need. Combination of Balstilimab/Zalifrelimab exhibit 26.5% objective response rates (ORR) vs 11.4% in Balstilimab monotherapy (102).

Microbiome
Microorganisms live on our skin, in our lungs, and in our intestines, accounting for a large portion of our cellular, metabolic, and genetic mass (103). The microbiota is the collective term for the species that make up this population, while the microbiome is the collective term for their genetic material. Inflammatory bowel disease, autism, asthma, and obesity have all been linked to the intimate relationship between microbiota composition, metabolic function, and immune system development and regulation (104). When a combination of broad-spectrum antibiotics (i.e., ampicillin plus colistin plus streptomycin) and single-agent imipenem elicited antitumor effects of CTLA-4 monoclonal antibody (mAb) as a result of microbiota impairment, the relationship between microbiota and ICIs was demonstrated. Antibiotic use between two months before and two months after the start of immunotherapy has been linked to a worse prognosis in patients treated with anti-PD-1/PD-L1 mAb (105). It is suggested that microbiota can influence dendritic cell (DC) maturation and activation. Several studies have reported the influence of bacteria such as Akkermansia mucinphila, Bacteriodetes, and Firmicutes in augmenting the response of ICIs. Creating a favorable microbiome using fecal transplantation in patients on ICIs is also being widely studied.
Anti-cytotoxic T-lymphocyte antigen 4 (anti-CTLA-4) immunotherapy causes mucosal damage and the translocation of Burkholderiales and Bacteroidales bacteria, which promote anti-commensal immunity as an adjuvant to anti-tumor immunity and are necessary for a positive response to therapy. Anti-PD-1/PD-L1 therapy, which does not harm the gut epithelia, requires pre-existing antitumor immunity, which is particularly effective in mice with intestinal Bifidobacterium spp (106) as portrayed in Figure 3.

Oncolytic Virus
Oncolytic viruses also appear to be an additional therapeutic agent for the management of cancer. These agents specifically target and kill cancer cells leaving the normal cells undamaged. Oncolytic viral infection triggers anti-cancer immune responses, which boost the effectiveness of checkpoint inhibition (107). Oncolytic viruses elicit anti-tumor immunity by promoting DC maturation and function. Improved DC maturation and function is because of the ability of oncolytic viruses in transferring genes encoding IFN-a, GM-CSF, and other cytokines. Breakdown of tumor cells induced by the oncolytic virus results in the release of DAMPS which further facilitates anti-tumor immune response. DAMPS include cell surface proteins, membrane proteins, and nucleic acid (108). FDA has approved an intralesional oncolytic immune therapy known as Talimogene Laherparepvec (T-VEC) for stage3b and stage 4 melanoma. A phase II trial comparing a combination of ipilimumab and T-VEC with ipilimumab monotherapy showed better ORR in the combination arm (39% vs. 18%; p < 0.02). Phase1b study of T-Vec combination with Pembrolizumab in melanoma patients reported no doselimiting toxicity with an ORR of 62% and a CRR (complete response rate) of 33% (109). Many studies including a phase III trial is underway (110).

Radiotherapy
It is known that radiation induces DNA damage-induced apoptosis and recent evidence suggests that it stimulates tumor antigen release and exerts an immune-mediated tumor response (120). Radiotherapy has been stated to enhance immunosuppressive TME via (1) Treg proliferation, M2 polarization of TAMs, and MDSC accumulation by increasing HIF 1 alpha transcription (2) activating latent TGF beta in the tumor, which converts CD4+ Tcells to Tregs and polarizes TAMs into M2 phenotypes (121,122). Some studies demonstrate the abscopal effect after radiation therapy distinctly from the treatment effects of immunotherapy alone (123 (127). GSK609 is an anti-Inducible T cell Co-Stimulator (ICOS) receptor agonist antibody that is used to treat cancers with various histologies. The first in man Phase 1 studies are ongoing as a single agent and along with Pembrolizumab. In patients with previously treated, PD-1/L1 naive HNSCC, preliminary evidence suggests that the combination of GSK609 and pembrolizumab has promising antitumor activity and a manageable safety profile (NCT02723955) (128).
Promising results are reported with anti-Her2 antibody (trastuzumab) in combination with ICIs in patients with HER2-positive metastatic esophagogastric cancer. A study reported a median PFS of 13·0 months (95% CI 8·6 to not reached), and a median OS of 27·3 months with a median DOR of 9.4 months (129). Phase 3 Keynote-811 trial is ongoing and expected to get a detailed result in combining pembrolizumab and trastuzumab (130). Similarly, clinical trials are ongoing to evaluate the response of other anti-tumor antibodies such as rituximab (anti-CD20) in combination with pembrolizumab in recurrent follicular lymphoma and DLBCL (NCT02446457) and cetuximab (anti EGFR)+ ICI combination in recurrent/ metastatic HNSCC (NCT03494322) and RAS wild type metastatic colorectal cancer (NCT04561336).
A very recent study published by Chan LC et al. found that JAK1, activated by IL-6 phosphorylates Tyr112 of programmed death-ligand 1 (PD-L1), causing STT3A, an endoplasmic reticulum-associated N-glycosyltransferase, to catalyze PD-L1 glycosylation and maintain PD-L1 stability. In animal models, targeting IL-6 with an IL-6 antibody resulted in synergistic T cell killing when paired with anti-T cell immunoglobulin mucin-3 (anti-Tim-3) therapy. In hepatocellular carcinoma patient tumor tissues, there was a positive association between IL-6 and PD-L1 expression. These findings point to the use of anti-IL-6 and anti-Tim-3 as a marker-guided therapeutic strategy (131).

Other Drugs in Combination With ICIs
There are a few drugs that are not used in cancer therapy however have been widely studied in combination with ICIs. The ultimate aim of using these drugs as adjuncts to ICIs is to create an immune stimulatory TME. Most of these studies are still in the preclinical setup but some of them are in clinical studies and show favorable outcomes ( Table 3).

Drugs Targeting Myeloid-Derived Suppressor Cells
The key targets in the TME are considered to be MDSCs. Research demonstrates that diminished immunotherapy efficacy is associated with the presence of MDSCs in the TME (132,133). MDSCs play a vital role in the immunosuppression of TME. These cells possess a strong immunosuppressive potential by representing a heterogeneous population of immature myeloid cells. Antitumor reactivity of NK cells and T cells have also been inhibited by MDSCs. They also encourage angiogenesis, the formation of pre-metastatic niches, and the recruitment of other immunosuppressive cells including regulatory T cells (134). Many commonly prescribed drugs such as rosiglitazone, cimetidine, tadalafil, etc. have shown preclinical benefits in targeting MDSCs (135). Even though the mechanism behind these drugs is different, the goal is to create an immune stimulatory TME inhibiting MDSCs. Examples of these classes of drugs with their proposed mechanism of action include: Paricalcitol a synthetic vitamin D analog has been used along with ICIs in some of the pancreatic clinical trials and showed a favorable response compared to conventional therapies. In a phase 2 trial using paricalcitol in combination with nivolumab, gemcitabine, nab-paclitaxel, and cisplatin in patients with untreated metastatic pancreatic ductal adenocarcinoma (PDAC), of the 25 patients enrolled thus far, OS has been recorded as 15·3 months, with an underlying ORR of 83% (NCT02754726).
A J Lugibuhl et al. conducted a study using Tadalafil Nivolumab combination in preoperative Head and Neck squamous cell cancer (HNSCC) showed 50% of patients having a pathologic treatment effect (TE) > 21% in 4 weeks. In the context of PD-1 blockade, tadalafil enhanced T lymphocyte and myeloid cell infiltration. 45 patients were treated with the combination regimen, and a preliminary assessment of pathologic treatment effect was performed on post-treatment specimens, revealing that 25% of patients had no treatment effect (TE), 25% had 1-20% TE, 41% had 21-99% TE, and 9% had CR (NCT03238365) (136).

Drugs Targeting Tregs
Regulatory T cells (Tregs) play a pivotal role in maintaining immune homeostasis and self-tolerance. They also play a negative role in inducing efficacious anti-tumor response. There is substantial evidence that depleting Tregs or inhibiting Treg function improves antitumor effects (137,138). The suppressive activity of Treg cells is exhibited by various mechanisms including secretion of inhibitory cytokines such as TGF beta, IL10, and IL35; inhibition of APC maturation through the CTLA-4 pathway; and expression of granzyme and perforin which kills effector T-cells (139,140). TGFb inhibitors and IDO inhibitors are currently explored along with checkpoint inhibitors in various cancer models such as melanoma, NSCLC, head, and neck, renal and bladder cancer, etc.
TGFb Inhibitors. Transforming growth factor (TGF) is a cytokine that induces Tregs and mediates immunosuppression in the TME. Therefore the strategy of blocking the recruitment of Tregs by TGFb inhibitors has shown benefits in some of the preclinical trials (141). T-cell sequestration away from the tumor mass is caused by the expression of a fibroblast TGF response signature (F-TBRS) in peritumoral fibroblasts. Immune checkpoint blockade with PD-1/PD-L1 inhibitors is ineffective in the absence of physical proximity between cytotoxic T cells and tumor cells. Pharmacological inhibition of TGFb reverses such immune exclusion and promotes CD8+ T cell penetration into tumors. TGF inhibition combined with immune checkpoint blockade maximizes tumor regression, with several tumor-bearing mice showing complete remission (142)(143)(144)(145). Active TGF-signaling in the peritumoral stroma, especially in patients with low tumorinfiltrating T cells in the tumor parenchyma, resulted in a lack of response to atezolizumab (anti-PD-L1 mAb) in metastatic urothelial cancer, according to a study (146). Galunisertib a TGFb inhibitor is in various clinical trials in combination with ICIs in  (147) and MDSCs, leading to suppression of the activity of T cells and NK cells (148). Several IDO inhibitors are being investigated in combination with CTLA-4 and PD-1 inhibition based on promising preclinical results. Epacadostat is an example of an IDO inhibitor that is being studied along with ICIs in clinical trials. COX-2 inhibitors which are another class of drugs have been found to have potential use as an adjunct with ICIs to treat immunosuppressive tumors that constitutively express indoleamine 2,3dioxygenase (IDO1) (149)(150)(151). In the absence of IDO activity, recent studies have discovered that Tryptophan 2, 3-dioxygenase (TDO), another essential enzyme in the kynurenine pathway, plays a compensatory role. As a result, a dual inhibitor of IDO and TDO is currently being investigated in preclinical studies to achieve maximum inhibition of the kynurenine pathway and alleviate tumor immune suppression (152).

Drugs Targeting the Renin-Angiotensin System (RAS)
RAS components are also expressed in many cell types of the TME, such as endothelial cells, fibroblasts, monocytes, macrophages, neutrophils, dendritic cells, and T cells (153,154). Growth and dissemination can be facilitated or hindered by RAS signaling in these cells which have been shown to affect cell proliferation, migration, invasion, metastasis, apoptosis, immunomodulation, angiogenesis, cancer-associated inflammation, and tumor fibrosis/ desmoplasia (155,156). The angiotensin II and angiotensin receptor (AngII/AT1R) axis regulates the tumor stroma and contributes to an immunosuppressive microenvironment through various mechanisms (Figure 4). Several clinical studies have also shown that RASi may have favorable effects in a wide range of malignancies. The gain in survival is tumor type-and stage-dependent and ranged from 3 months (advanced NSCLC) to more than 25 months (metastatic RCC) in retrospective studies (157). Drugs such as losartan are already in clinical trials as an adjunct with ICIs in many cancers.

Drugs Targeting the Autonomic Nervous System (ANS)
Lately, the role of the nervous system, in particular, the sympathetic nervous system has been observed in immune response regulation. Normally, the response of acute stress is beneficial whereas chronic stress is detrimental due to the suppression of effector immune cell activity while enhancing the immunosuppressive cell activity (158). Beta-blockers have been shown to have significant benefits when used along with ICIs. Data have shown that decreasing adrenergic stress by housing mice at thermoneutrality or treating mice housed with b-blockers at cooler temperatures reverses immunosuppression and remarkably increases responses to ICIs (159). Various preclinical, as well as clinical studies, have shown improved efficacy when ICIs are used along with a b blocker (159)(160)(161).

Metformin
Metformin has shown antitumor properties in various studies (162,163). Metformin prevents gluconeogenesis in the liver by regulating the adenosine monophosphate-activated protein kinase (AMPK)/liver kinase B1 (LKB1) pathway. The AMPK/ LKB1 pathway controls the cell cycle by regulating protein synthesis and cell proliferation by adjusting the amount of energy needed by the cells (164). Cancerous cells are inhibited and apoptosis is induced as a result of this regulation of the cell cycle and proliferation. Metformin was found to target CD8+ tumor-infiltrating leukocytes in a sample (TILs). Metformin also prevents the production of immune tolerance to cancer cells by inhibiting the synthesis of unfolded proteins, activating the immune response to cancer cells, inhibiting the expression of CD39/73 on MDSCs, and inhibiting the expression of CD39/73 on MDSCs. Metformin prevents exhaustion of CD8+ TILs, thus increasing production of various cytokines such as TNF-a, TNF-g, IL-2 and reversing immunosuppression. Some studies in NSCLC have reported that metformin exhibits cytotoxic effects as well (165)(166)(167). Preclinical and clinical studies are ongoing to explore the efficacy of metformin as an adjunct to ICIs. An interim safety analysis of Durvalumab Metformin combination in Head and Neck squamous cell carcinoma proved safe in part 1 of the study which contains only 6 patients and enrollment for part 2 with more patients is currently ongoing (NCT03618654) (168).

Future Implications
Various other modalities are being explored to achieve maximum benefit from ICI therapy. Newer checkpoint targets are being analyzed such as LAG-3, TIM-3, TIGIT, VISTA, or B7/H3. Moreover, stimulatory checkpoint pathway agonists such as GITR, 4-1BB, OX40, ICOS, CD40, or molecules targeting TME components like TLR or IDO are under investigation (169,170). Inhibiting multiple pathways along with ICIs is another ongoing topic of research. Combination of multiple checkpoint inhibitors and triplet or doublet therapy with adjunct drugs are being clinically validated. Improving TME to produce an immune stimulatory environment is being widely explored and drugs targeting these immune suppressive cells such as MDSC, TAM, etc. are being developed (Figures 5 and 6). Targeting mast cells is another approach that is currently being studied. Some of the recent studies have reported that intra-tumoral mast cell levels rose as the tumor progressed, and this independently predicted a shorter OS (171). It was found that CXCL12-CXCR4 chemotaxis results in the accumulation of these tumor-infiltrating mast cells. Intra-tumoral mast cells induced by cancer strongly express PD-L1 in both dose-dependent and time-dependent manner (172). This mechanism can be explored in the future to develop adjunct drugs with ICIs.

Translational Perspectives
As the spectrum of combinatorial immunotherapy strategies widens, the demand for robust preclinical models is at an alltime high. Combination strategies with ICIs require stalwart preclinical models for precise repurposing of the partner drugs.
Murine models generally used previously may not be helpful in immuno-oncology because of fundamental differences in the immune composition between mice and humans. The introduction of a patient-derived humanized xenograft model (Hu-PDX) is a very promising strategy in immuno-oncology. Hu-PDXs are mouse cancer models where tissue or cells from a patient's tumor are implanted into a humanized mouse which contains an immune system identical to humans (173). This knowledge of the Hu-PDX model further propelled research in immuno-oncology and enlightened the concept of co-clinical trial. In a co-clinical trial model, the clinical trial is coupled with a preclinical trial which provides valuable information to the corresponding clinical trial (174). Simultaneous treatment of the Hu-PDX model of a patient enrolled in a clinical trial with new agents helps in further understanding of underlying mechanisms, evaluating novel combination strategies, and identification of potential biomarkers. These advantages of coclinical trials will ensure the swift transition of preclinical data to the clinic which will be of great value to precision medicine and immuno-oncology.
Besides Hu-PDX models, patient-derived tumor threedimensional organoid models are also an exciting proposition for co-clinical trials (175). Organoids (and cancer-derived organoids) are three-dimensional tissue-like cellular clusters made from tissue or tumor-specific stem cells that imitate in vivo (tumor) characteristics such as cell heterogeneity (176). The major limitation is the inability to create a tumor-associated immune system in these models. Despite these limitations, organoid cultures accurately mimic the in vivo response of patient-derived xenograft models, as demonstrated in recent studies using bladder cancer-derived organoids by Pauli et al. and Lee et al. (177,178).

DISCUSSION
Our review shows that the majority of the adjuncts, which vary from chemotherapy, targeted therapy, radiation, immunemodulators, or even other forms of immunotherapy when combined with ICIs could provide far superior results than previously achieved. Adjunct drugs/modalities with favorable and unfavorable responses are distinctly illustrated in our review.
Maike Trommer, Sin Yuin Yeo et al. showed abscopal effects of RT in up to 29% of the patients when combined with anti PD1 immune check point inhibition (123). Huaqin Yuan et al. showed that Axitinib which is a known antiangiogenic agent could have an immune effect through its inhibition of MDSCs, STAT 3 pathway, and stimulation of CD8 T cells. This effect of Axitinib could enable us to use it as an adjunct along with checkpoint inhibition (179). A R Folgueras et al. showed that epigenetic modification could enhance immune surveillance in epithelial tumors. They found that HDAC inhibition could enhance the immunogenicity of epithelial tumors (180).
Similarly, David Roulois, Helen Loo Yau et al. found that hypomethylating agents may mount an antitumor response by inducing viral mimicry by endogenous transcripts (181). Recent studies have even used a triple-drug combination to augment antitumor immunity. Alexandra S. Zimmer, Erin Nichols et al. found a combination of PD1 inhibitor, Parp inhibitor, and antiangiogenic agent to be well tolerated and effective in a phase 1 study (75). Joseph A. Califano, Zubair Khan et al. found that even non-cancer drugs like tadalafil can have an immune-stimulating effect. Their work showed that tadalafil can reverse the resistance to immune recognition in head and neck cancer when patients were administered daily tadalafil (182).
Even though ICI's have shown tremendous promise especially in tumors refractory to standard treatments, their response rates are rarely beyond 15-20% (183). Translational research is of utmost value in this scenario and may help to further harness and improve the potential of immunotherapy.
Firstly, standard immune biomarkers like PDL1 levels, MSI status may not hold good when adjuncts are being used along with ICIs. Hence the need to develop newer biomarkers such as tumor mutational burden, tumor-infiltrating lymphocytes, miRNA studies, establishing 'foreignness' of neoantigens, HLA variations, N/L ratio, etc. so that a larger %age of patients can reap the benefits of ICIs. Artificial intelligence, neural networks, virtual clinical trials, and patient-derived xenograft models may be required to determine which patients may benefit from ICIs, to learn about the variation in the durability of response among patients, and identify those who may benefit from re-challenge of ICIs.
Secondly, the dose of adjuncts to be administered with ICIs would need to be determined. Ideally, an immune stimulatory' FIGURE 5 | Chemotherapy, Radiotherapy, and Epigenetic therapy which facilitates immunogenic cell death increase antigen presentation enhances cross-priming of dendritic cells, and decreases PDL-1 expression on tumor cells activates the antitumor immune response. Drugs that target Tregs, tumor-associated macrophages (TAM), and myeloid-derived suppressor cells block immunosuppression and enhance T effector T cell function to convert immunosuppressive tumor microenvironment into the immune-stimulatory environment.
dose is required which may or may not be the same as the standard dose used otherwise. Once proof of concept is established, the ICI-adjunct combination could be used in tumor agnostic/basket trials thus further tapping into unknown potentials of ICIs. Thirdly, although the mechanisms of ICI-adjunct combinations have been broadly defined, the exact additive/synergistic mechanisms need to be determined through well-constructed trials and translational research before they can become part of standard treatment. Fourthly, adverse effects of ICI-adjunct combination could be a lot greater than when either of the agents is used singly. Moreover, failure of combination strategies in the first-line may result in a lack of availability of drugs for second-line use. Hence proper analysis of adverse events and determining risk: benefit ratio would be required before such combinations become part of routine practice.
Finally, the emerging role of the microbiome and its effect on ICI combinations needs to be addressed and evaluated as it may play a major role in immune-modulation and effectiveness of such combinations. The above-addressed points all form a major portion of T0 and T1 phases of translational research which could greatly widen our understanding and applicability of ICIs and their combinations.

CONCLUSION
The efficacy of ICIs is highly variable and only a few patients are benefited from them. To address this issue there are many ongoing translational studies to target the resistant mechanisms of ICIs. The main principle is to enhance the TME to an immune stimulatory type from an immunesuppressive one. In this review, we have evaluated available literature which included drugs used as adjuncts in clinical trials along with ICIs. Many drugs have shown significant benefits and some have shown a mixture of both favorable and unfavorable outcomes. There is conflicting evidence for most of the above-mentioned adjuncts such as metformin, Vit D, etc. From the available adjuncts, the selection of the best adjunct for the patients on ICI treatment is still a dilemma. Not only the selection of adjuncts but also the dose for the desired property is also questionable. It might require a dose that is different than used normally and needs to be further evaluated. Therefore, the need of conducting co-clinical trials or translational studies in this field is immense. We believe that recent advances in preclinical models such as humanized xenograft models, invitro modeling using 3D bioprinting, etc. may give further insight into this problem.

AUTHOR CONTRIBUTIONS
HV: Data collection, scientific writing, proof reading, and concept. VS: Scientific writing, proof reading, and concept. BT: Data collection and scientific writing. SM: Concept, proof reading, and scientific writing. RN: Proof reading and project overseer. All authors contributed to the article and approved the submitted version.