Checkpoint Inhibition: Will Combination with Radiotherapy and Nanoparticle-Mediated Delivery Improve Efficacy?

Checkpoint inhibition (CPI) has been a rare success story in the field of cancer immunotherapy. Knowledge gleaned from preclinical studies and patients that do not respond to these therapies suggest that the presence of tumor-infiltrating lymphocytes and establishment of immunostimulatory conditions, prior to CPI treatment, are required for efficacy of CPI. To this end, radiation therapy (RT) has been shown to promote immunogenic cell-death-mediated tumor-antigen release, increase infiltration and cross-priming of T cells, and decreasing immunosuppressive milieu in the tumor microenvironment, hence allowing CPI to take effect. Preclinical and clinical studies evaluating the combination of RT with CPI have been shown to overcome the resistance to either therapy alone. Additionally, nanoparticle and liposome-mediated delivery of checkpoint inhibitors has been shown to overcome toxicities and improve therapeutic efficacy, providing a rationale for clinical investigations of nanoparticle, microparticle, and liposomal delivery of checkpoint inhibitors. In this review, we summarize the preclinical and clinical studies of combined RT and CPI therapies in various cancers, and review findings from studies that evaluated nanoparticle and liposomal delivery of checkpoint inhibitors for cancer treatments.


Introduction
The emergence of immunotherapy into the mainstream of oncology has been fueled by recent clinical advances and FDA approvals of inhibitors that block immune checkpoints in cancers, such as non-small-cell lung cancer, melanoma, head and neck squamous cell carcinoma, pancreatic ductal adenocarcinoma, lymphoma, and renal cell cancers [1][2][3]. Immune checkpoint molecules, such as programmed death receptor 1 (PD-1), programmed death-ligand 1 (PD-L1), cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), T-cell immunoglobulin, and mucin domain-containing-3 (TIM-3), and lymphocyte-activation gene 3 (LAG-3), are expressed in various activated immune cells, tumor cells, and other tissues, and lead to inhibition of immune responses against tumors [4]. Two of the most studied checkpoint receptor/ligand interactions include PD-1/PD-L1 and CTLA-4/CD28. Intracellular signaling emanating from these interactions leads to suppression of effector immune responses against the tumors. Upon receptor/ligand interaction, the cytoplasmic tail of the receptor gets tyrosine phosphorylated on the tyrosine residue containing regions known as immunoreceptor tyrosine-based switch motifs (ITSMs) [5]. This leads to the recruitment of Src-homology domain-containing phosphatase 2 (SHP-2). The phosphatase activity of SHP-2 results

Preclinical Studies of Radiation and Checkpoint Inhibitor Immunotherapy
Preclinical studies in various cancer models have demonstrated that the combination of RT and CPI therapy, such as anti-CTLA-4 and anti-PD-L1, can enhance antitumor immune responses and improve survival [11][12][13][14][15]. These preclinical studies, however, have not yet achieved a consensus on the best sequence of administration of these combination therapies or the RT dose and fractionation that would optimize therapeutic efficacies. There are many variables, both within the TME and surrounding stroma of tumors, that influence the results of these combination strategies. Hence, it is unlikely that there will be one standard treatment plan for all tumor types when combining CPI with RT (CPI-RT). The following section summarizes published preclinical studies that help build the knowledge and understanding of the complex interactions and effects of CPI-RT that could help informed design of future clinical trials.
When RT was combined with both PD-L1 and CTLA-4 blockade, further improvements in antitumor responses, complete responses (CRs), and survival proportions, were achieved in preclinical models. This synergy of anti-CTLA-4 or anti-PD-L1 with a range of RT doses and fractionations has been demonstrated in immune-competent mouse models of lung, breast, melanoma, and colorectal cancers [11,12,[14][15][16][17][18]. Studies have shown that CPI enhances therapy responses in RT-resistant tumors that overexpress checkpoint molecules in response to RT. RT, on the other hand, improves antitumor responses in poorly immunogenic tumors that did not respond well to CPI therapies by enhancing T-cell receptor repertoire and immunogenic cell death, leading to tumor-antigen release [11,12,16,18]. In a mouse model of colorectal cancer (CT26 cell line), RT-mediated local control was significantly improved (p < 0.001), with concurrent anti-PD-L1 or anti-PD-1; resulting in curative rates of 66% and RT and/or CPI can also improve therapy responses by modulating the immune-suppressive cells, such as myeloid-derived suppressor cells (MDSCs) and regulatory T cells (Tregs), present in the TME. MDSCs are a heterogeneous population of immature cells of myeloid origin. These cells regulate immune responses by suppressing the T-cell responses. High dose (30 Gy) alone has been shown to decrease the frequency of MDSCs (unirradiated vs. irradiated = 26% vs. 6%) in tumors, while resulting in the increased infiltration of CD8 + T cells (unirradiated vs. irradiated = 19% vs. 70%); a process that was dependent on CD8 + antigen cross-presenting DCs, IFNγ secretion, and CD40Lexpressing CD4 + T cells [30]. However, in this study, an extended fractionated regimen was found to be inefficient in controlling metastases or in enhancing survival. In another preclinical study, ablative hypofractionated RT, but not conventional fractionated RT, was shown to decrease the infiltration of MDSCs in the tumors through a mechanism dependent on inhibition of vascular endothelial growth factor (VEGF) production [20]. CTLA-4 and/or PD-1/PD-L1 blockade, in preclinical and clinical studies, has been shown to reduce MDSCs in the TME of various tumors [12,[31][32][33][34]. The mechanisms of decreased MDSCs in these studies include decreased levels of CXCL1, inhibition of the CD47/SIRPα pathway, or antagonism by increased release of the T helper cell 1 (Th-1) type of cytokines. Combining RT with CPI (such as anti-PD-L1) can further decrease the MDSCs in TME (RT vs. RT plus anti-PD-L1 = 4.78% ± 2.49% vs. 0.38% ± 0.16% of total CD45 + cells), resulting in enhanced antitumor responses [12]. While RT reduces MDSCs, it increases the production and recruitment of regulatory T cells (Tregs) in the TME [35]. CPI, on the other hand, decreases the infiltration of Tregs [33]. PD-1 blockade has been shown to decrease FoxP3 expression in a preclinical model [36], and Tregs from patients that responded to PD-1 blockade had diminished suppressive function [37]. While RT and/or CPI can also improve therapy responses by modulating the immune-suppressive cells, such as myeloid-derived suppressor cells (MDSCs) and regulatory T cells (T regs ), present in the TME. MDSCs are a heterogeneous population of immature cells of myeloid origin. These cells regulate immune responses by suppressing the T-cell responses. High dose (30 Gy) alone has been shown to decrease the frequency of MDSCs (unirradiated vs. irradiated = 26% vs. 6%) in tumors, while resulting in the increased infiltration of CD8 + T cells (unirradiated vs. irradiated = 19% vs. 70%); a process that was dependent on CD8 + antigen cross-presenting DCs, IFNγ secretion, and CD40L-expressing CD4 + T cells [30]. However, in this study, an extended fractionated regimen was found to be inefficient in controlling metastases or in enhancing survival. In another preclinical study, ablative hypofractionated RT, but not conventional fractionated RT, was shown to decrease the infiltration of MDSCs in the tumors through a mechanism dependent on inhibition of vascular endothelial growth factor (VEGF) production [20]. CTLA-4 and/or PD-1/PD-L1 blockade, in preclinical and clinical studies, has been shown to reduce MDSCs in the TME of various tumors [12,[31][32][33][34]. The mechanisms of decreased MDSCs in these studies include decreased levels of CXCL1, inhibition of the CD47/SIRPα pathway, or antagonism by increased release of the T helper cell 1 (Th-1) type of cytokines. Combining RT with CPI (such as anti-PD-L1) can further decrease the MDSCs in TME (RT vs. RT plus anti-PD-L1 = 4.78% ± 2.49% vs. 0.38% ± 0.16% of total CD45 + cells), resulting in enhanced antitumor responses [12]. While RT reduces MDSCs, it increases the production and recruitment of regulatory T cells (T regs ) in the TME [35]. CPI, on the other hand, decreases the infiltration of T regs [33]. PD-1 blockade has been shown to decrease FoxP3 expression in a preclinical model [36], and T regs from patients that responded to PD-1 blockade had diminished suppressive function [37]. While CTLA-4 blockade has been shown to selectively deplete T regs in a Fc-Fcγ-dependent manner in mouse models [38], it is unable to deplete T regs in human cancers [39], highlighting the need and potential for modifications of Fc portions for enhanced Fc-mediated depletion of T regs . RT-CPI decreases T regs in the TME [40][41][42], suggesting that combined RT-CPI may overcome inefficacies of either treatment alone, by decreasing the immunosuppressive cells.
The effectiveness of RT and its combination with CPI may also depend on the dosing strategy. In a study, an RT dose of 11.5 Gy × 2 was able to inhibit VEGF receptor signaling, and lead to subsequent reduction in MDSCs, yet 4 Gy × 9 fractions did not have the same effect [20]. Also, while higher doses of RT can cause more dsDNA damage and promote cell death, RT doses above 12-18 Gy, per fraction, can also induce TREX1, which degrades cytoplasmic dsDNA, the component needed to trigger the STING pathway, resulting in decreased IFN-β and, subsequently, reduced abscopal effects. An RT dose between 8-12 Gy can result in the highest frequency of dsDNA breaks before triggering TREX1 elevation. Repeated doses of RT that do not trigger TREX1 would lead to increased IFN-β production, recruitment, and activation of dendritic cells, and subsequent priming of T cells for improved antitumor responses [26]. Therefore, multiple fractions of 8-12 Gy may be the ideal RT regimen to achieve a balance between immune-inhibitory and immunostimulatory signals in the TME for optimal antitumor responses [20,22,25,26]. Further studies, however, are necessary to determine the optimal dose(s) and fractionations that may convert each tumor type into an in situ vaccine for reliable antitumor responses.
In addition to the doses, timing and sequencing of RT and CPI are also crucial to the success of combined RT-CPI treatment. Combined RT-CPI has been shown to be effective if given concomitantly [11,19], but not sequentially (PD-1/PD-L1 blockade after RT) [11], with the efficacy of the concomitant treatment attributed to acute increase in PD-1 expression on the infiltrating T cells. The RT-induced surge in antigen presentation, TILs, and PD-1/PD-L1, only lasts for a couple days after RT; therefore, anti-PD-1/PD-L1 should be given ideally within 3-5 days of RT, if RT is the first treatment in the sequence [11,19]. Similarly, in a preclinical model of breast cancer, RT followed at least a day later by administration of anti-CTLA-4, improved survival compared to either treatment as monotherapy [16]. Pretreatment with anti-CTLA-4 followed by RT, however, was not evaluated. Alternatively, in a mouse model of colorectal cancer, treatment with anti-CTLA-4 one week prior to treatment with RT, was shown to be superior to treatment with anti-CTLA-4 one week after the RT [43]. This effect was attributed partly to the anti-CTLA-4-mediated depletion of T regs . Efficacy of stimulation of the co-stimulatory molecule OX40, however, relied on administration of agonist antibody at least a day after the RT [43]. Agonist OX40 antibody has also been shown to help overcome resistance to anti-PD-1/PD-L1 [44,45]. Anti-OX40 can boost tumor-specific T-cells in non-immunogenic mouse models prior to anti-PD-1/PD-L1 administration, resulting in improved local and distant tumor control and survival [43][44][45]. Timing of anti-OX40 is important to harness the efficacies of the combination treatments. If anti-OX40 is being used to enhance anti-PD-1/PD-L1 activity, it should be given several days prior to anti-PD-1/PD-L1 therapy [45]; however, if anti-OX40 is being used to enhance RT-effects, then it should be given immediately after RT to coincide with and harness the RT-induced antigen release and subsequent T-cell activation [43]. In the case of RT combination with a checkpoint blockade (CTLA-4) and checkpoint stimulation (OX40), Young et al. showed that administration of anti-CTLA-4 prior to RT, followed by OX40 stimulation, was the optimal combination in a mouse model of colorectal cancer [43]. These results highlight the complexities and need for further studies in determining the optimal sequencing of RT with inhibitors and agonists against different checkpoint and co-stimulatory molecules, respectively.
The last 20 years of preclinical research has helped to fill a void of understanding of the mechanisms of responses and resistances to combined RT-CPI in mouse models of cancers. Informed by such studies, various RT dosing and sequencing, in combination with CPI, are being evaluated in clinical trials. It will be interesting to see if these mechanisms, hypotheses, and expected outcomes are confirmed in these ongoing and future clinical trials with RT-CPI combination.

Clinical Trials of Radiation and Checkpoint Inhibitor Immunotherapy
Ipilimumab (Bristol-Myers Squibb), a fully human CTLA-4 monoclonal antibody, was approved by the FDA in 2011 [46]. Monotherapy demonstrated improved long-term survival in subsets of patients with advanced melanoma [47,48]. An increased response rate has been demonstrated with 10 mg/kg, compared to lower dosing, without unacceptable toxicity [49], although non-response is still prevalent. Efforts have been made to enhance efficacy in a larger patient subset with the addition of other immune-modulating therapies. Ipilimumab, in combination with PD-1/PD-L1 blockade, has improved response rates and survival, albeit with increased toxicity and cost [49]. The application of immunotherapy has been broadened with data supporting use in cancers of the lung, kidney, and head and neck, with expansion of indications underway [50][51][52][53]. Tables 1-3, list the current clinical trials under investigation.
Preclinical, retrospective, and Phase 1 data using hypofractionated and/or stereotactic body radiation therapy (SBRT) combined with ipilimumab and/or PD-1/PD-L1 blockade have suggested synergism without added toxicity [54][55][56][57][58][59][60]. There is a suggestion that improved tumor responses may occur with higher radiation doses, and when delivered in close proximity to immunotherapy [61]. The current decade has witnessed an exponential increase in clinical research investigating combined immunotherapy and radiation therapy [62]. The earliest and most robust data have been presented for advanced melanoma, although there is emerging evidence in other solid and hematologic malignancies as well. The following section summarizes the results of completed and ongoing clinical trials evaluating the efficacy of RT-CPI on patients with various tumors.

Melanoma
The earliest prospective experience with combined CPI and RT was reported in 2015 from the University of Pennsylvania [18]. Twenty-two patients with metastatic melanoma were enrolled in a phase I trial in which a single index lesion received hypofractionated irradiation, followed by four cycles of ipilimumab. Among the 12 patients that were evaluated by PET for the irradiated lesion, none had progressive metabolic disease [18]. Of the unirradiated lesions, 18% of patients experienced a partial response (PR), 18% had stable disease (SD), and 64% had progressive disease (PD). Median progression-free survival (PFS) and overall survival (OS) was 3.8 and 10.7 months, respectively [18]. A subsequent phase I trial recruited 22 patients who received radiotherapy (both hypofractionated and standard fractionation) to one to two sites within 5 days of starting ipilimumab [55]. Fifty percent demonstrated clinical benefit, with 27.3% achieving ongoing complete response (CR) at median f/u of 55 weeks, and 27.3% achieving a PR for median of 40 weeks [55]. Patients who achieved a CR tended to have a smaller volume of disease and baseline, and experienced higher grade hypophysitis, in line with prior reports demonstrating improved control among patients who experience more significant immune-related toxicity [63].
Boutros et al. reported a phase 1 SBRT dose escalation trial in combination with ipilimumab (10 mg/kg for 4 doses) in 19 patients with advanced melanoma [54]. Radiotherapy was administered in 9, 15, 18, and 24 Gy in 3 fractions. Maximum tolerated dose (MTD) of 9 Gy was demonstrated, as two of six patients receiving 15 Gy experienced dose-limiting toxicity (DLT). The objective response rate (ORR) was 21%, with four patients experiencing PR and another four experiencing SD. The median PFS and OS were 7.2 and 4.4 months, respectively [54]. A similar trial was reported recently from Belgium [64]. Twelve patients with metastatic melanoma were enrolled in a phase 1 trial of dose-escalated SBRT (24 Gy, 30 Gy, and 36 Gy in 3 fractions) to one lesion and 4 cycles of ipilimumab at 3 mg/kg. SBRT was delivered before the third cycle of immunotherapy (IT). Local control was achieved in all but one irradiated patient, and the maximum tolerated dose (MTD) was not reached. Three patients experienced abscopal response in non-irradiated lesions. Grade 3-4 IT-related toxicity occurred in 25% of patients [64].
Given the high incidence of brain metastases in melanoma patients and poor intracranial response to ipilimumab alone, early combination experience with whole brain radiotherapy (WBRT) or stereotactic radiosurgery (SRS) has been reported to optimize intracranial control [65][66][67]. Efficacy and safety of combined SRS with PD-1 blockade was reported in retrospective single institution reports [68]. There is suggestion that the presence of radionecrosis is associated with prolonged OS and improved disease control [69]. Williams et al. reported a phase 1 trial of 16 patients treated with combined ipilimumab and either WBRT or SRS, depending on the degree of intracranial disease burden [70]. WBRT was delivered as 30 Gy in 10 fractions, and SRS was based on maximum tumor diameter or size of resection cavity, according to dose prescriptions on RTOG 90-05 trial [71]. Ipilimumab was started at 3 mg/kg on day 3 of WBRT, or 2 days after SRS, with an independent escalation of dose to 10 mg/kg. No patients experienced dose-limiting toxicity or radionecrosis. In contrast to the historical median of 4.7 months in melanoma patients with brain metastases, median OS was 8 months in the WBRT arm, and not reached in the SRS arm [72].

Central Nervous System
The efficacy of combined IT and RT for brain metastases is largely comprised of melanoma data, and has been described above. There is a lack of evidence supporting combination therapy for primary brain tumors. Keynote-028 demonstrated efficacy of pembrolizumab (anti-PD-1, Merck, Kenilworth, NJ, USA) in 26 PD-L1-positive recurrent glioblastoma-multiforme patients with a median OS of 14 months and median PFS of 3 months with a low rate of toxicity [73]. Multiple prospective trials are currently enrolling patients with high grade gliomas investigating the combination of RT and IT, such as CPI.

Head and Neck
Patients who experience recurrence or metastases from a head and neck primary tumor often have a poor prognosis and limited therapeutic options. Nearly 40% of pathologic specimens demonstrate the presence of tumor infiltrating lymphocytes, providing a rationale for the efficacy of CPI [74]. Keynote-012 enrolled 60 patients in a phase Ib trial who received pembrolizumab at 10 mg/kg [74]. An overall response was seen in 18%, including 25% in HPV-positive patients and 14% in HPV-negative patients, and the drug was well tolerated [74]. Keynote-040 randomized 495 patients with recurrent squamous cell carcinoma of the oral cavity, oropharynx, hypopharynx, or larynx to pembrolizumab or investigator choice of standard doses of methotrexate, docetaxel, or cetuximab [53]. Although there was a higher overall response rate with pembrolizumab, there was no statistical difference in OS or PFS, albeit with lower grade (3)(4)(5) adverse events [53]. There is no current published prospective data on RT plus checkpoint blockade for the treatment of head and neck cancer.

Non-Small Cell Lung Cancer (NSCLC)
Keynote-024 compared pembrolizumab vs. investigator's choice of cytotoxic chemotherapy in 305 patients with an advanced NSCLC and PD-L1 tumor proportion score of ≥50% [75]. This phase 3 trial demonstrated the superiority of pembrolizumab, compared to platinum-based chemotherapy, with results showing an increased median PFS, increased overall survival at 6 months, and increased median duration of response with less treatment-related adverse events [75]. There is no prospective evidence supporting the addition of RT to anti-PD-1/PD-L1 therapy in advanced disease, with a number of ongoing trials.
Thoracic SBRT regimens with biologically effective doses (BED) of approximately 100 Gy have been shown to have improve local disease control [76]. This dose is higher than that delivered with conventional radiation, and raises concerns about safety in combination with CPI. A phase 1 trial conducted at MD Anderson Cancer Center investigated concurrent or sequential SBRT to lung or liver lesions in a dose-escalated fashion combined with ipilimumab at 3 mg/kg [77]. Concurrent or sequential 50 Gy in 4 fractions or sequential 60 Gy in 10 fractions was prescribed to 35 patients. Response outside the radiation field, the primary response metric, demonstrated 10% partial response and 23% experienced clinical benefit (PR or SD lasting ≥6 months). Two patients receiving liver SBRT experienced DLT, one receiving 50 Gy concurrently, and the other receiving 50 Gy sequentially. There were no DLTs in the lung patients. Thirty-four percent experienced grade 3 toxicity, and no patients experienced grade 4-5 adverse effects [77]. A phase II study conducted in the Netherlands reported on 72 patients with advanced NSCLC randomized between of pembrolizumab alone or of pembrolizumab preceded by SBRT (8 Gy × 3 within 7 days) [78]. ORR was doubled (19% vs. 41%), and median PFS was tripled (1.8 vs. 6.4 months) with the addition of SBRT, demonstrating that SBRT augments the antitumor immune response [78].

Small Cell Lung Cancer (SCLC)
SCLC is a highly aggressive malignancy, with 70% presenting with late stage disease. Nearly all patients experience local and/or distant progression, and no studied therapy to date has demonstrated an improvement over the standard of care (platinum-based chemotherapy) [79]. CheckMate 032 demonstrated durable efficacy and safety of nivolumab (anti-PD-1, Bristol-Myers Squibb) monotherapy, and in combination with ipilimumab, in a phase 1/2 trial [80]. A phase 3 trial evaluated the efficacy and safety of standard of care chemotherapy with or without ipilimumab in 954 patients with newly diagnosed extensive-stage small cell lung cancer [81]. Unfortunately, there was no difference in median OS or PFS with a higher rate of treatment-related discontinuation in the combination treatment group [81]. There are no reported prospective data investigating CPI with RT in SCLC.

Breast
Breast cancer represents a spectrum of disease genotypes; among which, the triple negative and Her2-postive subtypes have been found to be immunogenic [82,83]. Early data on CPI monotherapy in locally advanced/metastatic breast cancer has demonstrated a modest benefit. Keynote-086 demonstrated an overall response rate of 5%, with median duration of response of 6.3 months in a subset of patients with heavily pretreated metastatic triple negative disease [84]. Ongoing studies are evaluating CPI in combination with standard cytotoxic chemotherapy. The addition of RT to pembrolizumab was assessed in a single arm phase II study of 17 patients, unselected for PD-L1 expression, with metastatic triple negative disease [85]. A dose of 30 Gy was delivered in 5 fractions of 6 Gy, within 3 days of pembrolizumab infusion. Of 9 evaluable women, 33% had PR, 11% SD, and 56% had PD with no added toxicities [85]. Several further trials are ongoing.

Gastrointestinal
The role for CPI in locally advanced/metastatic esophageal, gastroesophageal junction (GEJ), and gastric cancers, is emerging. Keynote-028, a phase Ib trial, reported outcomes in 23 patients with advanced esophageal/GEJ tumors treated with pembrolizumab [86]. ORR was 30.4%, with 13% SD, and 12-month PFS of 21.7% with manageable toxicities [86]. Preliminary data from the phase Ib, Keynote-012 trial in 39 patients with gastric cancer treated with pembrolizumab, noted a 22% overall response, with a manageable toxicity profile [87]. There are no prospective data, to date, reporting on combined CPI and RT, although multiple trials are underway.
There have been limited advances in the management of pancreatic adenocarcinoma. Patients typically present with locally advanced or metastatic disease. Concomitant chemoradiation extends median survival from 4.1 to 6.1 months, and gemcitabine administration adds another mere 1.24 months of survival. A single arm phase II study explored ipilimumab in 27 patients and demonstrated no response in all but one patient, who experienced delayed regression of the primary lesion and hepatic metastases [88]. To date, there have been no published prospective data on RT plus CPI for the treatment of pancreatic cancer. Similarly, there is limited data supporting the use of IT in hepatobiliary cancers or small/large bowel malignancies.

Genitourinary
There are several treatments approved for the treatment of metastatic castrate resistant prostate cancer after progression with docetaxel chemotherapy, all of which have been demonstrated to improve OS compared to control [89]. Pathologic specimens often demonstrate inflammatory cell infiltrates, suggesting a host immune response. A phase I/II ipilimumab (10 mg/kg) dose escalation study, in combination with 8 Gy of RT to a bone lesion in 84 patients, reported efficacy with tolerable adverse effects [90]. Of 50 patients that received this dose, 8 had PSA declines of ≥50%, one had a CR, and six had SD [90]. CA184-043 was a multicenter phase 3 trial of men with metastatic castrate-resistant prostate cancer, who experienced progression after docetaxel chemotherapy [89]. A total of 799 patients received either bone-directed radiotherapy (8 Gy × 1) followed by ipilimumab (10 mg/kg) or placebo. Median OS was 11.2 vs. 10.0 months (p = 0.053), with an increase in toxicity among the patients receiving ipilimumab [89].
Renal cell carcinoma (RCC), that has disseminated, has limited therapeutic options that provide moderate overall survival benefit. Nivolumab was compared to everolimus in a phase 3 study of 821 pretreated patients with advanced clear-cell carcinoma, and demonstrated an OS of 25.0 vs. 19.6 months (p = 0.002), with a 25% ORR compared to 5% and lower grade 3-4 toxicities [51]. Ipilimumab was subsequently added to nivolumab, and compared to sunitinib in 1096 previously untreated patients [91]. The 18-month OS was 75% vs. 60%, and median survival was not reached, vs. 26.0 months. ORR was 42% vs. 27%, with lower grade 3-4 toxicities [91]. SBRT has been increasingly used in the management of inoperable primary RCC or management of metastatic disease, with overall local control of 85-100% [92]. There are no reported prospective data on combined IT/CPI and RT, and trials are currently underway.
The use of intravesicular BCG in 1976 first demonstrated the efficacy of immunotherapy in urothelial carcinoma of the bladder [93]. Platinum-based chemotherapy, however, has been the standard of care for advanced diseases with limited overall survival benefit. A multicenter phase II trial reported on 310 patients with inoperable locally advanced or metastatic urothelial carcinoma, with progressive disease after platinum-based chemotherapy [94]. Patients received atezolizumab (anti-PD-L1, Roche, Indianapolis, IN, USA) and demonstrated an ORR of 26% with an ongoing response in 84% at median follow-up of 11.7 months, with 16% developing grade 3-4 adverse effects [94]. There are no reported prospective data of combined RT with CPI for urothelial carcinoma, and trials are underway.

Gynecologic
Treatment options for recurrent and/or metastatic cancers of the cervix, uterus, vagina, and vulva, are limited after first-line therapy. CheckMate-358, a phase I/II study, reported preliminary results in 24 women with cancer of the cervix, vagina, and vulva, treated with nivolumab, and demonstrated an ORR of 20.8% and median PFS of 5.5 months [95]. Keynote-028 reported data on 28 women with locally advanced or metastatic PD-L1-positive endometrial cancer, with progressive disease after standard therapy. The patients were treated with pembrolizumab at 10 mg/kg every two weeks. Thirteen percent achieved PR, and 13% SD with median duration of response not reached, and with no patients experiencing grade 4 adverse events [96]. Although RT is commonly used in the management of primary and recurrent malignancies of the gynecologic tract, there are no reported data combining CPI with RT, and trials are ongoing.

Nanoparticle Delivery of Checkpoint Inhibitors
The primary goal of cancer immunotherapy is to stimulate the host immune system to help eliminate cancer cells [97]. While CPI therapies embody this goal, they are often costly, delivered systemically, and may be discontinued in patients who have severe immune-related toxicities [98]. In this regard, nanoparticle delivery vehicles may overcome some of these barriers by improving stability and delivery of checkpoint inhibitors to tumor sites. Nanoparticles are particles that have a size in the range of nanometers. Different types of nanoparticles include, but are not limited to, liposomes, dendrimers, metal nanoparticles, carbon nanoparticles, silica nanoparticles, and magnetic nanoparticles. Various modifications of these nanoparticle platforms are often used to facilitate passive or targeted delivery of therapeutic and imaging agents to the tumor tissues. Design and composition of an ideal nanoparticle incorporates desired characteristics, such as biodegradability, ease of fabrication, cost-effectiveness, non-immunogenicity, and enhanced permeation and retention, with sustained release of payload at the tumor site [99]. Enhancing the therapeutic index of drug molecules is a major rationale of nanoparticle drug delivery systems, and modalities that incorporate CPIs with nanoparticle for therapies will not only have potential to improve therapeutic efficacies by enhanced delivery, but will also limit systemic toxicities [98]. In this regard, studies have explored the pharmacodynamics/pharmacokinetics, as well as therapeutic efficacies of CPIs' incorporation into nanoparticle delivery vehicles for cancer therapies.

Polymeric and Metal Nanoparticle Delivery of Checkpoint Inhibitors
Metal-core nanoparticles and polymer nanoparticles have been studied for their efficacy in incorporating and delivering checkpoint inhibitors to tumor sites. A reporter polymeric nanoparticle, carrying paclitaxel, and which incorporated PD-L1-blocking antibodies through conjugation with PEG, showed enhanced antitumor activity in preclinical models of lung and breast cancer, leading to significantly decreased tumor volumes (p < 0.001) compared to control nanoparticles [100]. In another study, iron-dextran nanoparticles were conjugated with blocking antibody against PD-L1 and agonistic antibody against the co-stimulator 4-1BB [101]. This allowed for simultaneous blockade of checkpoint molecule, PD-L1, and stimulation of co-stimulatory molecule, 4-1BB, resulting in robust activation of tumor-infiltrating CD8 + T cells (increased CD107 + and IFNγ + CD8 + T cells; p < 0.05), decreased average tumor size, and improved survival in preclinical models of melanoma and colon cancers. [101]. In melanoma model, the tumor sizes for antibody-conjugated nanoparticles (ACN) vs. no treatment were 112 mm 2 vs. 205 mm 2 , respectively (p < 0.001) [101]. A significant decrease (p < 0.01) in tumor size was also observed with can, as compared to free antibody injections. Similarly, for colon cancer model, the tumor sizes were 19 mm 2 vs. 158 mm 2 (p < 0.001), for ACN vs. no treatment, respectively [101]. Animal survival in the colon cancer model was significantly increased from 10% for untreated mice to 70% (p < 0.001) for ACN-treated mice [101]. This study also determined that the in vivo half-life of ACN was 84.5 h, compared to 15.2 h for soluble antibody (p < 0.0001), with retention of 60% ACN as compared to 8% for soluble antibody at 72 h post-injection [101]. These studies highlight the potential for improved therapeutic efficacies and decreased toxicities, due to nanoparticle-mediated delivery of chemotherapeutic drugs, as well as immunomodulatory antibodies, such as checkpoint inhibitors, to the tumor site.

Liposomal Delivery of Checkpoint Inhibitors
Liposomes have been used as vehicles for chemotherapeutic drug delivery to the tumors. Liposomes are versatile nanoparticles that can be tailored for precision medicine. Multiple preclinical and clinical investigations, evaluating the use of nanoparticles and liposomes for delivering antibodies, genes etc. to the tumor sites, have emerged in recent years [98,[102][103][104][105]. Liposomes are spherical lipid vesicles that are comprised of an aqueous core encapsulated by one or more lipid bilayers [106,107]. Modifications in preparation methods allow for generation of liposomal particles with different structures, colloidal size, surface charge, and chemical compositions as well as conjugations [106]. These design flexibilities can be exploited to create liposomes that can overcome barriers in drug delivery and imaging. In addition to chemotherapeutic drugs, modifications of liposomal delivery vehicles also permit attachment of different therapeutic and targeting antibodies, enabling targeted and sustained delivery to the tumor site. Characteristics of liposomes, such as biocompatibility, modulated pharmacokinetics, enhanced bioavailability, etc., make liposomes a promising delivery system for various drugs, genes, and immune therapies [106], and have led to preclinical and clinical investigations of the feasibility and efficacy of liposomes as therapeutic and diagnostic tools.
While liposomes have many advantages, one of the drawbacks of using conventional liposomes as drug carriers is their susceptibility to rapid clearance by the reticuloendothelial system. Scientists have tried to overcome this barrier by PEGylating the liposomes. PEGylation involves conjugation of the liposomal particles with polyethylene glycol (PEG). This increases size, and creates a protective hydrophilic layer on the surface of liposomes, resulting in decreased clearance by the reticuloendothelial system and kidneys [106,108]. The advantages of PEGylation include decreased immunogenicity, extended circulation time, enhanced pharmacokinetic profile, and improved drug solubility and stability [109]. In a recent study, efficacy of doxorubicin-loaded liposomes that were conjugated to DSPE-PEG-PD-1 monoclonal antibody, was evaluated. The results showed significant tumor growth inhibition (p < 0.05) with PD-1-conjugated liposomes compared to irrelevant IgG-conjugated liposomes [110]. These conjugated liposomes were also found to be stable for at least 48 h when incubated in serum, suggesting stability in biological systems [110]. A similar preclinical study with PEGylated liposomes carrying CTLA-4 blocking antibody showed improved accumulation into the tumor (PEGylated vs. non PEGylated vs. free anti-CTLA-4 = 7.57 + 1.55% ID/g vs. 0.63 + 0.43% ID/g vs. 1.06 + 0.42% ID/g respectively; p < 0.01; ID/g = injected dose per gram of tissue) and half-life, resulting in significant tumor growth delay (PEGylated vs. non PEGylated vs. free anti-CTLA-4 = 29.37% vs. −2.07% vs. 17.57% respectively) and improved median survival (PEGylated vs. non PEGylated vs. free anti-CTLA-4 = 34.98 vs. 22.27 vs. 30.12 days respectively; p = 0.0001) compared to non-PEGylated formulation or CTLA-4 antibody treatment alone [111]. Efficacy of PEGylated liposomes in delivering antibodies to the tumor site was confirmed in yet another study, with results also showing enhanced stability of liposomes and prolonged preservation of the secondary and tertiary structures of the delivered antibodies [112]. PEGylation of liposomes has been shown to reduce immunogenicity, and diminish complement activation and clearance by immune system [113,114], hence making them an attractive delivery vehicles for immunomodulatory antibodies, such as checkpoint inhibitors. Another liposomal formulation, nanohybrid liposomal cerasome nanoparticles (NLCNPs), was evaluated in a separate study [115]. Compared to non-conjugated PD-L1 administration along with paclitaxel, NLCNPs, carrying paclitaxel and conjugated with anti-PD-L1 antibodies, was significantly more efficient in delivering the drugs to the tumor site, resulting in enhanced tumor control and inhibition of metastases without added toxicities [115]. These studies underscore the benefit of using various formulations/modifications of liposomes for prolonged half-life and targeted delivery of checkpoint modulatory antibodies, without affecting their structure and function, to the tumor site for enhanced therapeutic efficacy.

Microparticles Delivery of Checkpoint Inhibitors
Silica and poly (lactic-co-hydroxymethyl-glycolic acid)-based microparticles have also been evaluated for delivery of CPIs and resulting therapeutic efficacies and toxicities [116,117]. Rahimian et al. showed that sustained release (up to 80% release in 30 days) of immunomodulatory antibodies at the tumor site, over time, can be achieved by intratumoral injection of antibodies loaded microparticles [116]. The microparticles were based on biodegradable poly (lactic-cohydroxymethyl-glycolic acid) (pLHMGA), and were loaded with blocking antibody to CTLA-4, and agonistic antibody to CD40. Although the therapeutic efficacy of the antibody-loaded microparticles was similar to the control formulation (antibodies with incomplete Freund's adjuvant (IFA)), significantly lower amounts of antibodies (5-10 times lower compared to antibodies in IFA) were detected in the serum of the microparticle formulation-treated animals, suggesting that this may lead to decreased systemic toxicities [116]. Similarly, in a mouse model of melanoma, intratumoral injection of functionalized mesoporous silica-based microparticles (with pore size up to 30 nm in diameter), that allowed for sustained release of anti-CTLA-4 antibody, slowed tumor growth (p < 0.05), and improved survival, compared to systemic administration of CTLA-4 blocking antibody or IgG conjugated microparticles [117]. Comparison with direct intratumoral injection of unconjugated anti-CTLA4 antibody, however, was not made in this study. These studies emphasize the potential for sustained release of CPIs at tumor sites, and decreased toxicities upon microparticle-mediated delivery.
While some clinical trials (such as NCT02158520 and NCT03107182) are evaluating the efficacy of nanoparticle delivery of chemotherapies in combination with systemic PD-1 or CTLA-4 blockade with or without RT, to the best of our knowledge, targeted deliveries of checkpoint inhibitors by nanoparticles or microparticles, with or without RT, have yet to be studied. It remains to be seen if the targeted delivery of CPIs using nanoparticles or microparticles enhances the therapeutic efficacy in combination with RT, and what schedule and dose combinations derive the best clinical outcomes.

Conclusions
Advances in clinical and preclinical sciences have shown that both CPI and RT have vast potentials in controlling and treating cancer malignancies. The clinical outcomes, however, are limited, due to innate or therapy-induced adaptive resistances that undermine the efficacy of RT or CPI as stand-alone treatments. Understanding the underlying mechanisms of resistance has led to studies aimed at evaluating the combination of RT and CPI. While results have been promising, these studies also highlight the need to further evaluate the sequence of treatments, doses, and fractionation schedules, and the type of checkpoint molecules targeted, in combination with RT, in order to generate the optimal therapeutic responses. Additionally, durability of the RT-CPI-generated T-cell responses, and determinants of abscopal responses, remains to be fully understood. Focused preclinical studies and ongoing clinical trials of RT-CPI should answer some of these outstanding questions, and aid in determination of optimal sequencing, dosing/fractionation, and selection of the RT-CPI treatments for specific tumor types.
One of the major roadblocks to successful CPI therapies against cancers includes the immune-related toxicities associated with systemic CPI treatments. In this regard, targeted delivery of checkpoint inhibitors has the potential to overcome this barrier. Various formulations of nanoparticles, liposomes, and microparticles, have been studied, to determine their feasibility as vehicles to deliver and provide sustained release of checkpoint molecules to the tumor site. Preclinical studies have shown that targeted delivery of checkpoint inhibitors not only enhances the efficacy of the treatments, but also decreases toxicities. While many clinical studies have evaluated antitumor efficacy of targeted delivery of chemotherapies and immunotherapies to tumor site, no clinical studies evaluating the nanoparticle/liposome/microparticle-mediated delivery of checkpoint molecules to TME are available. Additionally, nanoparticle-and microparticle-mediated delivery of CPI, in combination with RT, represents another opportunity to generate optimal antitumor responses with decreased toxicities. Further studies are warranted, however, to determine if such a combination has a clinical rationale.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.