Combining vaccines and immune checkpoint inhibitors to prime, expand, and facilitate effective tumor immunotherapy

Abstract

Introduction:

Multiple immune checkpoint inhibitors (ICIs) that modulate immune cells in the periphery and the tumor microenvironment (TME) have been approved, as have the therapeutic cancer vaccines sipuleucel-T for metastatic castration-resistant prostate cancer and talimogene laherparepvec (T-VEC) for metastatic melanoma. These developments provide rationale for combining these modalities to improve response rates and durability of responses in a variety of cancers. Preclinical data have shown that vaccines can induce immune responses that turn a tumor from ‘cold’ to ‘hot,’ but vaccines do not appear to be highly active as monotherapy.

Areas covered:

Here, we provide a review of the current state of vaccine and ICI combination studies.

Expert commentary:

Most combination trials are in early phases, but several are now in phase III. Vaccines that target antigens expressed exclusively on tumor cells, neoantigens, have the potential to induce robust antitumor responses. Several techniques for predicting which neoepitopes to target, based on tumor mutational profiling, are in various stages of development. To be successful, combination immunotherapy approaches must seek to prime the immune system, expand the immune response, and facilitate immune function within the TME.

1. Introduction

In recent years, monoclonal antibodies that antagonize immune inhibitory receptors, known as immune checkpoint inhibitors (ICIs), have transformed the treatment landscapes of several malignancies [1–7]. Blocking the inhibitory functions of these molecules can affect interactions among several different types of immune cells in a variety of immune compartments, including peripheral lymph nodes and the tumor microenvironment (TME). Generally speaking, this dampens inhibitory pressures on the immune system and produces antitumor activity in some patients. The most impressive response rates have been observed in tumors with immune infiltrates, i.e., ‘hot’ tumors with clear underlying immune recognition such as melanoma and non-small cell lung cancer (NSCLC) [3,8–10]. Nonetheless, response rates with ICIs alone leave much room for improvement. For example, in advanced melanoma, objective response rates (ORRs) for ICIs range from 19% with ipilimumab, an antibody against cytotoxic T lymphocyte-associated protein 4 (CTLA-4), to 43.7% with nivolumab, an antibody against programmed cell death protein 1 (PD-1), and 57.6% with the combination [9]. Although a large number of preclinical studies combining therapeutic cancer vaccines and ICIs have been completed or are ongoing, only a handful of clinical trials have been conducted to date.

Combination immunotherapy approaches that prime the immune system, expand the immune response, and facilitate immune function within the TME can improve the clinical activity of ICIs (Figure 1). Therapeutic vaccines that engage the immune response by generating antigen-specific T cells are a key component of this prime, expand, and facilitate strategy.

Figure 1.

Multimodal immunotherapy can prime, expand, and facilitate the anti-tumor immune response. Tumor-targeted vaccines can generate antigen-specific T cell responses. Immune checkpoint inhibition can enhance those responses by blocking negative regulation during antigen presentation in peripheral lymphoid organs. Cytokines can expand and activate tumor-specific T cells as well as NK cells. Within the tumor microenvironment (TME), blocking immunosuppressive pathways can enable the anti-tumor immune response. For example, PD-1/PD-L1 blockade at the tumor:effector cell synapse can enhance tumor lysis. TGF-neutralization, when combined with PD-L1 inhibition, has been implicated as means to enable T cell infiltration into tumor. IDO-1 enzyme inhibition and immune agonists can also splay the balance of immune suppression toward a more immune permissive state within the TME by dampening inhibitory effects of myeloid-derived suppressor cells (MDSCs). Increased tumor cell death leads to release of tumor antigens and may lead to cross priming of additional antigen-specific T cell responses.

Two targeted vaccines are currently approved by the U.S. Food and Drug Administration (FDA): sipuleucel-T for metastatic castration-resistant prostate cancer (mCRPC) and talimogene laherparepvec (T-VEC) for unresectable stage IIIB, IIIC, or IV melanoma. Many therapeutic cancer vaccines that employ various platforms and target a myriad of tumors are in development. After expansion of vaccine-derived, tumor-directed T cells, ICIs can facilitate antitumor activity within the TME and elsewhere. The most extensively studied ICIs are those targeting CTLA-4, PD-1, and programmed death-ligand 1 (PD-L1). These molecules and their ligands are largely involved in regulating T cells during antigen presentation and in modulating regulatory T cells (Tregs) and the lytic function of CD8 + T cells. The immune functions of these molecules have been reviewed elsewhere [11–15]. Here, we review clinical reports of vaccine/ICI combinations and discuss promising avenues under exploration.

2. Immune checkpoints

CTLA-4 is expressed exclusively on T cells and binds to its ligands, CD80 and CD86, on antigen-presenting cells. While a complete understanding of CTLA-4 biology is currently lacking, in preclinical models, these interactions have been observed to have distinct effects on different T cell subsets. For example, during antigen presentation, CTLA-4 is upregulated following activation of naive T cells. CTLA-4 competes with the activating CD28 receptor that is also expressed on T cells for binding to CD80 and CD86 expressed on antigen-presenting cells. This interaction not only inhibits expansion of memory T cells, but can also favor downstream generation of Tregs rather than CD4+ helper T cells [15]. Further characterization of the relationship between CTLA-4 and Tregs is an area of active investigation [16].

The PD-1 receptor is expressed on activated effector T cells and Tregs [17,18]. PD-1 binds to PD-L1 and PD-L2 on cancer cells, and primarily checks peripheral autoimmunity and promotion of self-tolerance [19]. PD-L1 expression promotes immune evasion by tumor cells within the TME and can be upregulated by IFN made by activated tumor-targeting T cells [20,21].

3. Fda-approved therapeutic cancer vaccines

A comprehensive discussion of cancer vaccine platforms is beyond the scope of this review and has been published elsewhere [22]. Several vaccine platforms have been developed, including cellular, oncolytic, peptide, viral, and bacterial. Delivery methods include intramuscular, intravenous, intratumoral, and subcutaneous administration. There are currently two FDA-approved therapeutic cancer vaccines: sipuleucel-T and T-VEC.

The first FDA-approved therapeutic cancer vaccine was sipuleucel-T, an infusional cellular product generated by incubating a patient’s own peripheral blood mononuclear cells (PBMCs) with the prostate tumor-associated antigen prostatic acid phosphatase (PAP) and granulocyte-macrophage colony-stimulating factor (GM-CSF). Since dendritic cells (DCs) derived from PBMCs process PAP and generate antigen-specific T cells, sipuleucel-T is considered a DC vaccine. Data demonstrate an increase PAP-specific T cells [23], as well as an increase in T cell infiltration of prostate tumor following vaccination [24].

Sipuleucel-T was FDA-approved in 2010 for men with asymptomatic or minimally symptomatic mCRPC following results of the phase III IMPACT trial (n = 512). Patients who received sipuleucel-T every 2 weeks for a total of three doses had a median overall survival (OS) of 25.8 months versus 21.7 months for placebo (HR: 0.78; 95% CI: 0.61–0.98; p = 0.03). There were no statistically significant differences in time to disease progression between trial arms [25]. Consensus guidelines suggest using this agent earlier in mCRPC in patients without aggressive disease characteristics [26]. Antigen spreading appeared to be associated with improved OS [27,28].

T-VEC is a first-in-class intralesional oncolytic viral vaccine that was FDA-approved in 2015 for the local treatment of unresectable cutaneous, subcutaneous, and nodal lesions in patients with recurrent melanoma. It is composed of a modified herpes simplex virus type 1 encoding GM-CSF. T-VEC has a twofold mechanism of action: replication within neoplastic cells leads to lysis of the cancer cells, which releases tumor-associated antigens (TAAs) and primes antitumor T cell responses. T-VEC also contains two copies of the GM-CSF gene that attract and stimulate DCs, leading to cross-priming and generation of antigen-specific T cell responses [29–33].

In the phase III OPTiM trial (n = 436) in unresectable stage III or IV melanoma, T-VEC produced an improved durable response rate (16.3% vs. 2.1%), overall response rate (26.4% vs. 5.7%) and complete response rate (11% vs. 1%), with a trend toward improved OS (23.3 months vs. 18.9 months; p = 0.51) compared with GM-CSF alone [34]. T-VEC has also been shown to decrease the size of 34% of uninjected non-visceral lesions and 15% of visceral lesions, demonstrating its ability to generate systemic immunity leading to antitumor activity distant from the site of injection [35].

4. Fda-approved vaccine/ici combinations

4.1. Ipilimumab plus T-VEC in melanoma

A phase Ib/II study evaluated the safety and efficacy of T-VEC in combination with ipilimumab in 198 patients with unresectable stage IIIB-IV melanoma. This was the first randomized trial to evaluate an oncolytic viral agent in combination with an ICI. A first dose of T-VEC was given at ≤ 4 mL × 10⁶ plaque-forming units (pfu) per mL; after 3 weeks it was given at ≤ 4 mL × 10⁸ pfu/mL every 2 weeks. Ipilimumab 3 mg/kg was administered intravenously every 3 weeks for up to four doses starting at week 1 in the ipilimumab-alone arm and in week 6 in the combination arm. Totally, 38 patients (39%) in the combination arm and 18 patients (18%) in the ipilimumab-alone arm had an objective response (odds ratio: 2.9; 95% CI: 1.5–5.5; p = 0.002). Responses were not limited to injected lesions; decreases in the size of visceral lesions were observed in 52% of patients in the combination arm and 23% of patients in the ipilimumab-alone arm. The study showed that the combination of T-VEC plus ipilimumab had greater antitumor activity without additional toxicity compared with ipilimumab alone [36].

4.2. Pembrolizumab plus T-VEC in melanoma

T-VEC was also evaluated in combination with pembrolizumab, an anti-PD-1 antibody, in a similar melanoma population in the phase Ib portion of the phase Ib/III clinical trial called MASTERKEY-265. In total, 21 patients with untreated, unresectable stage IIIB-IV melanoma were treated with T-VEC in (sub)cutaneous/nodal lesions at a first dose of 4 mL × 10⁶ pfu/mL and, starting 3 weeks later, with 4 mL × 10⁸ pfu/mL every 2 weeks. Pembrolizumab 200 mg was given intravenously starting with the third dose of T-VEC and then every 2 weeks. Per immune-related response criteria (irRC), the confirmed ORR was 62% and the complete response (CR) rate was 33%. In patients who responded, there was also an increase in the number of CD8 + T cells, increased tumoral PD-L1 protein expression, and interferon (IFN)-γ gene expression after the start of T-VEC therapy. Response did not appear to be associated with baseline CD8 + T cell infiltrate or baseline IFN-γ levels [37]. The ORR in this study was approximately double as seen in a phase III trial of pembrolizumab (34%) and in the T-VEC OPTiM trial (26%). The findings of this study suggest that the combination of T-VEC plus an ICI modulates the TME and thus improves treatment efficacy. The randomized, double-blind phase III KEYNOTE-034 trial of T-VEC plus pembrolizumab versus T-VEC placebo plus pembrolizumab is ongoing (NCT02263508).

4.3. Pembrolizumab plus T-VEC in head and neck squamous cell carcinoma

Data from a phase Ib/III multicenter, open-label, randomized trial of T-VEC plus pembrolizumab for recurrent or metastatic head and neck squamous cell carcinoma (HNSCC) were presented at the 2017 Congress of the European Society for Medical Oncology [38]. At that time, 28 patients who had progressed after receiving platinum-based therapy had enrolled in the safety phase of the trial. One patient died of an arterial hemorrhage after two doses of both drugs. Grades 3 and 4 adverse events were seen in six patients (38%), but none led to treatment discontinuation. There were five severe adverse events (chills, pyrexia, stridor, odynophagia, and arterial hemorrhage) possibly related to T-VEC, and two (eczema and pyrexia) related to pembrolizumab. The combination regimen was deemed safe and is now in the efficacy portion of the trial [38].

4.4. Ipilimumab plus sipuleucel-t in prostate cancer

A phase I study of sipuleucel-T plus escalating doses of ipilimumab enrolled nine patients with progressive mCRPC. Patients received three doses of sipuleucel-T at a minimum dose of 50 million autologous CD54+ cells activated with PAP fusion protein (PA2024) every 2 weeks, followed 1 week later by ipilimumab 1 mg/kg intravenously. Ipilimumab was given to cohorts of three patients either 1 week after the third dose of sipuleucel-T, 1 and 4 weeks after, or 1, 4, and 7 weeks after. Statistically significant increases in serum antibodies specific for PAP and PA2024 were seen after sipuleucel-T treatment and increased again significantly post-ipilimumab treatment. Cancer-specific immunoglobulin titers began to decline after 5–11 months. There were no unexpected toxicities from sipuleucel-T and ipilimumab [39]. As of abstract publication in February 2018, six patients are alive and the median survival has surpassed 4 years. One patient remains in durable remission [40]. Significant increases in immune correlates and promising survival data were observed in this small trial despite the low dose of ipilimumab, opening the door for larger trials with higher doses of an ICI.

There are currently six T-VEC plus ICI and two sipuleucel-T plus ICI trials actively enrolling patients (as of 18 July 2018; see clinicaltrials.gov).

4.5. Ipilimumab plus vaccine

In 2011 ipilimumab, an anti-CTLA-4 antibody was awarded FDA approval for the treatment of previously treated unresectable or metastatic melanoma, based on a phase III randomized (3:1:1), double-blind, double-dummy trial (n = 676); the vaccine plus ICI arm was negative. Patients were randomly assigned to receive either ipilimumab in combination with a glycoprotein 100 (gp100) peptide vaccine (n = 403), ipilimumab plus vaccine placebo (n = 137), or vaccine plus placebo (n = 136). Ipilimumab 3 mg/kg was administered intravenously with or without vaccine every 3 weeks for up to four treatments. The vaccine included two modified HLA-A*0201-restricted peptides injected subcutaneously as an emulsion with incomplete Freund’s adjuvant (IFA). No difference in OS was seen between the ipilimumab plus vaccine arm and the ipilimumab-only arm (10 months vs. 10.1 months; p = 0.76), and results for both ICI arms were better than the vaccine-only arm (6.4 months). Patients treated with ipilimumab alone had the best ORR (investigator assessed) of 10.9% compared with 5.7% in the combination arm and 1.5% in the vaccine-only arm [1].

PROSTVAC (PSA-TRICOM) is a PSA-targeted vaccine platform consisting of a recombinant vaccinia-derived prime dose followed by fowlpox-derived boost doses that also incorporate three costimulatory molecules (B7.1, ICAM-1, and LFA-3). Ipilimumab plus PROSTVAC and GM-CSF were tested in a phase I trial in 30 patients with mCRPC. Ipilimumab was given intravenously at doses of 1, 3, 5, and 10 mg/kg starting 2 weeks after a prime dose of PROSTVAC and was continued monthly on the same day as vaccine for six doses, then every 3 months if disease was stable. The combination of vaccine plus ipilimumab did not exacerbate the immune-related adverse events associated with ipilimumab, and 58% of patients had a decline in PSA. Six patients had a PSA decline > 50%. Median OS with the combination was 31.6 months, which compares well to data from the phase II trial of PROSTVAC (OS 25.1 months in the PROSTVAC arm vs. 16.6 months in the control arm) enrolled in the same treatment era [41]. There were trends for increased OS in patients with a longer predicted survival per Halabi nomogram, longer baseline PSA doubling time, and higher baseline hemoglobin level. Immune subset analysis showed an association with longer OS in several immune cell subsets before immunotherapy. An increase in mature natural killer (NK) cells post-vaccination was also associated with increased OS [42].

4.6. Nivolumab plus peptide vaccine

Prior to the 2014 FDA approval of nivolumab for advanced melanoma, several combination studies with nivolumab plus vaccines were conducted based on nivolumab’s clinical activity in melanoma and renal cell carcinoma. A phase I trial of 30 HLA A*0201+ patients with previously treated stage IV melanoma evaluated a multipeptide (MART-1/gp100/NY-ESO-1 with Montanide ISA 51 VG) vaccine plus intravenous extended-dose nivolumab (1, 3, or 10 mg/kg) every 2 weeks for 24 weeks, followed by nivolumab alone every 3 months. Data from this trial were presented at the 2012 Annual Meeting of the American Society of Clinical Oncology. Patients had not been previously treated with ipilimumab. Of 12 evaluable patients, three had confirmed partial responses and two had unconfirmed partial responses. At 12 weeks, patients in all cohorts showed a decrease in PD-1 expression on peripheral CD4+ and CD8 + T cells (p < 0.0001), increased CD4+ CTLA-4 levels (p = 0.01), and dose-related decreases in CD8 + T cells, whereas CD4 + T cells increased (both p < 0.05). The combination was well tolerated at 1 and 3 mg/kg [43], and the trial was expanded to include ipilimumab-refractory patients. A total of 90 patients with unresectable stages III or IV melanoma who were either ipilimumab-naive and had experienced progression after at least one prior therapy (cohorts 1 to 3, 34 patients) or had experienced progression after prior ipilimumab (cohorts 4 to 6, 56 patients) received nivolumab at 1, 3, or 10 mg/kg intravenously every 2 weeks for 24 weeks, then every 12 weeks for up to 2 years, with or without the multipeptide vaccine. RECIST response rates were 25% for both ipilimumab-refractory and ipilimumab-naive patients. Negative PD-L1 staining did not rule out a response. Circulating Tregs decreased in responders and those with stable disease and significantly increased in nonresponders at 12 weeks. In patients with ipilimumab-refractory or ipilimumab-naive melanoma, nivolumab at 3 mg/kg with or without peptide vaccine was well tolerated and induced responses lasting up to 140 weeks. Vaccination did not add to the clinical activity of nivolumab [44]. The same combination of multipeptide vaccine plus extended doses of nivolumab (1, 3, 10 mg/kg) was tested in another study of 33 patients with resected stage IIIC and IV melanoma; 31 patients had stage IV disease. In previous prospective [45,46] and retrospective [47] trials of resected stage IV melanoma, OS ranged from 12–21 months, whereas in this combination study, at a median follow-up of 32.1 months median OS was not reached and 23 patients remained disease-free. Again, PD-L1 tumor status was not significantly associated with relapse-free survival, and there was a trend toward lower baseline circulating Tregs and myeloid-derived suppressor cells in nonrelapsing patients [48].

4.7. Pembrolizumab plus viral vaccine

Pembrolizumab in conjunction with vaccine was evaluated in a phase I disease-agnostic setting, targeting p53 mutations in a variety of solid tumors that had failed or were intolerant to standard therapy. Patients received genetically engineered modified vaccinia virus Ankara (MVA) expressing a wild-type p53 transgene (p53MVA) at doses of 5.6 × 10⁸ pfu for three doses in combination with pembrolizumab 200 mg intravenously for seven doses every 3 weeks. Two patients (triple negative breast cancer and HNSCC) showed clinical benefit associated with durable p53-specific CD8 + T cell responses. The breast cancer patient had evident dermal metastases over > 50% of her body; genomic profiling showed no expression of PD-L1 on tumor cells prior to study treatment. After two cycles of therapy, her cutaneous metastases began to regress, and skin biopsy showed complete pathologic response in one lesion. By week 9, her skin lesions had regressed almost completely. Systemic response was also confirmed with CT and bone scans. Peripheral blood showed persistent activation of p53-specific CD8 + T cell responses that were associated with lymphocytic infiltration of the resolved dermal metastases. Clinical response lasted for more than 6 months after the initiation of combined therapy [49,50].

Several case studies of pembrolizumab plus vaccine have reported dramatic and unexpected results in cancers with poor prognosis. For example, a patient with uterine cervical small cell carcinoma, a rare and aggressive subtype of cervical cancer for which there is no standard therapy, was treated with pembrolizumab 1 mg/kg intravenously and an autologous formalin-fixed tumor vaccine including TAAs after she had progressed on chemotherapy with cisplatin and etoposide. She received three doses of vaccine 1–2 weeks apart, followed by pembrolizumab every 3 weeks. CT scan prior to the third dose of pembrolizumab revealed a liver lesion that had enlarged to 4.2 cm since the previous CT done after the last dose of vaccine. Totally, 11 days after the third dose of pembrolizumab, CT scan was repeated and the liver lesion had decreased to 1.9 cm [51]. The patients with uterine carcinoma and triple negative breast cancer, both heavily pretreated terminal and aggressive cancers without further standard treatment options, responded to the combination of vaccine plus anti-PD-1 therapy.

4.8. Anti-pd-l1 antibody plus vaccine

To our knowledge, no clinical trials combining anti-PD-L1 therapy and vaccine have reported results. Currently, many trials combining vaccines and anti-PD-L1 inhibitors (Table 1) are actively recruiting patients in a range of solid and hematologic malignancies (as of 17 March 2018; see clinicaltrials.gov).

Table 1.

Selected ongoing combination trials of vaccine plus anti-PD-L1 antibody.

Phase	Patient population	Therapy	1° Endpoint	NCT
Ib	TNBC (stage III/IV)	Adjuvant PVX-410 (tetra-peptide XBP1 TF-targeted vaccine) + durvalumab	Safety	02826434
I/II	Platinum- resistant/refractory peritoneal malignancies	Sequential and concurrent ONCOS-102 (intraperitoneal viral oncolytic vaccine) + durvalumab	Safety	02963831
II	AML after chemotherapy-induced remission	DC/AML vaccine (fusion of autologous AML cells and DCs); DC/AML vaccine + durvalumab; observation	PFS	03059485
Ib/IIa	HPV+ R/M HNSCC	MEDI0457 (HPV DNA vaccine) + durvalumab	Safety and ORR	03162224
I/II	Metastatic NSCLC	BI 1,361,849 (mRNA vaccine encoding 6 NSCLC-associated antigens) + durvalumab; BI 1,361,849 + durvalumab + tremelimumab	Safety	03164772
I/II	Refractory colon cancer	Pexa-Vec (viral oncolytic vaccine); Pexa-Vec + durvalumab; Pexa-Vec + durvalumab + tremelimumab	Safety	03206073
II	Advanced gynecologic cancers	Vigil (autologous tumor cell vaccine) followed by Vigil + atezolizumab; atezolizumab followed by Vigil + atezolizumab	Safety	03073525
Ia/Ib	Locally advanced or metastatic tumors	RO7198457 (mRNA vaccine targeting personalized TAAs) + atezolizumab	Safety	03289962
I	Advanced solid tumors	DSP-7888 (peptide vaccine derived from WT1 protein) + atezolizumab; DSP-7888 + nivolumab	Safety and RPIID	03311334
I	mCRPC	Sipuleucel-T followed by atezolizumab; atezolizumab followed by sipuleucel-T	Safety	03024216
I/IIb	Recurrent ovarian, fallopian tube, or primary peritoneal cancer	Atezolizumab; atezolizumab + guadecitabine; atezolizumab + guadecitabine + CDX-1401 (DC vaccine targeting NY-ESO-1)	Safety and PFS	03050814
II	Previously untreated metastatic CRC	FOLFOX + bevacizumab followed by maintenance bevacizumab + capecitabine; FOLFOX + bevacizumab + avelumab + Ad-CEA (adenovirus CEA-targeted vaccine) followed by maintenance bevacizumab + capecitabine + avelumab + Ad-CEA	PFS	03050814
I/II	R/R HPV 16+ cancers; R/M oropharyngeal cancer	TG4001 (viral vector HPV-targeted vaccine) + avelumab	Safety and ORR	03260023
II	Biochemically recurrent prostate cancer	PROSTVAC (viral PSA-targeted vaccine) + CV301 (viral CEA- and MUC1-targeted vaccine) + M7824 (anti-PD-L1 and anti-TGF-β agent)	Decline in PSA	03315871
I/II	Pancreatic cancer after progression on standard of care	ETBX-011 (adenovirus CEA-targeted vaccine) + GI-4000 (yeast Ras-targeted vaccine) + aldoxorubicin + high-affinity NK cells + avelumab + bevacizumab + capecitabine + cyclophosphamide + ALT-803 (IL-15 super agonist) + 5-FU + leucovorin + nab-paclitaxel + omega-3-acid ethyl esters + oxaliplatin + SBRT	Safety and ORR	03387098

5-FU: 5-fluorouracil; AML: acute myeloid leukemia; CEA: carcinoembryonic antigen; CRC: colorectal cancer; DC: dendritic cell; FOLFOX: 5-fluorouracil, leucovorin, oxaliplatin; HNSCC: head and neck squamous cell carcinoma; HPV: human papilloma virus; IL: interleukin; mCRPC: metastatic castration-resistant prostate cancer; mRNA: messenger RNA; MUC1: mucin 1; NK: natural killer; NSCLC: non-small cell lung cancer; ORR: objective response rate; PD-L1: programmed death-ligand 1; PFS: progression-free survival; PSA: prostate-specific antigen; R/M: recurrent/metastatic; RPIID: recommended phase II dose; R/R: recurrent/refractory; SBRT: stereotactic body radiation therapy; TAA: tumor-associated antigen; TF: transcription factor; TNBC: triple negative breast cancer; WT1: Wilms tumor gene 1.

5. Conclusion

To date, vaccine/ICI combinations have produced incremental improvements in response rates and occasional durable CRs, but no combination has proven universally, or even predictably, effective in all tumors of a certain subtype. Even when vaccines have been shown to induce a robust immune response, clinical response does not always follow. Many combination studies are early-phase, single-arm trials, making it difficult to accurately compare results of combinations versus ICI or vaccine monotherapy, but some combinations appear to lead to improved outcomes. Importantly, the combination of ICI and vaccine does not appear to increase the toxicity profile above what is observed when the agents are used as monotherapy. We are just beginning to explore potential immunotherapy combinations, including the rational combination of vaccine and ICI, and results will likely continue to improve.

6. Expert commentary

The success of a therapeutic cancer vaccine lies in its ability to induce an effective immune response at sites distant from the vaccination site. Unfortunately, the majority of vaccine monotherapy studies have produced modest or negative results, suggesting additional agents are needed. For example, sipuleucel-T, FDA-approved for mCRPC, produces a modest survival benefit of approximately 4 months, but does not typically induce PSA declines or objective responses [25]. T-VEC, FDA-approved for advanced unresectable melanoma, is also associated with a relatively small improvement in survival and response rates [35]. Since many cancer patients are heavily pretreated, resulting in compromised immune function, optimization of the immune system is important and may be achieved with an prime, expand, and facilitate approach.

Various tumor vaccines have demonstrated the ability to prime the immune system and induce immune responses specific to antigens contained within the vaccine and others that are not, a phenomenon known as antigen spreading or antigen cascade. Antigen spreading can occur with both subcutaneous and intratumoral vaccine administration [52,53]. Several studies have even shown stronger immune responses to cascade antigens than target antigens expressed by the vaccine [54–56]. The antigen spreading phenomenon, as well as tumor flares or pseudoprogression that can occur with immunotherapy, can make it difficult to accurately assess response in clinical trial patients and highlights the importance of incorporating irRC into these trials.

Antigen target selection is another important aspect of vaccine design. Therapeutic cancer vaccines can target antigens expressed exclusively on tumor (tumor-specific antigens) or antigens that are highly expressed on tumor but also expressed to some degree in normal tissue (TAAs). Many antigens exist somewhere on the spectrum between TAAs and self-antigens, with unclear distinctions in some cases. Targeting these antigens can lead to increased side effects and a less robust clinical response.

In response, neoantigen targeting has emerged as a promising strategy for vaccine development. Neoantigens are derived from somatic DNA alterations in tumors and, therefore, are highly tumor-specific. Next-generation sequencing of DNA has improved methods for identifying somatic DNA mutations. However, not all DNA mutations arise in coding regions and not all coded mutations are ultimately translated into peptide. Furthermore, in order for a neoantigen to affect antitumor immunity, the mutant peptide must be processed and presented by major histocompatibility complex into an epitope with adequate T cell receptor binding affinity [57]. Methods for predicting which immunologically relevant neoepitopes will be generated by a given mutation have been reported [58], and clinical validation of these techniques could result in personalized cancer vaccines. Trials testing neoantigen vaccines have been previously reviewed and there are plans for combination studies with immune checkpoint inhibition [59].

The choice of cancer vaccine adjuvants may also dictate synergy with ICIs by determining the site of immune response. The landmark ipilimumab trial in melanoma revealed that adding vaccination with a gp100 peptide formulated in IFA [1], a water-in-oil emulsion of antigen in mineral oil with mannide monooleate as a surfactant, lacked clinical benefit [60]. Preclinical studies have shown that vaccines emulsified with IFA produce depots that persist at the site of injection, which may lead to sequestration of tumor-specific CD8 + T cells at the vaccination site rather than localization to the tumor. T cells may become dysfunctional and eventually undergo apoptosis. Vaccination with short-lived formulations such as water-soluble peptides, viral vectors, or DCs allows for T cell localization to tumors and synergism with ICI [61,62].

The timing of vaccine administration in relation to checkpoint blockade also likely plays a role in clinical response, but ideal timing may differ among different vaccine/ICI combinations. Preclinical studies have shown that checkpoint receptors change in a time-dependent fashion after vaccine administration. In a murine model, CTLA-4 blockade was most effective in controlling tumor growth when administered one day after treatment with an engineered oncolytic vaccinia virus, whereas PD-1 blockade was most effective when delivered 7 days after vaccine treatment. These results correspond with findings that expression of CTLA-4 is significantly downregulated in activated T cells after 7 days, while PD-1 shows increased expression for a longer period [63]. Another murine study showed that administration of anti-CTLA-4 antibody and oncolytic vaccine on the same day did not result in antitumor benefit [64]. Another study administered anti-CTLA-4 and anti-PD-1 antibodies 6 days prior to a DC vaccine in an effort to mimic what investigators felt would occur in clinical practice, and produced a significant reduction in the rate of tumor progression and improvement in long-term survival [65]. The vaccine in this latter study is currently under investigation in a phase I metastatic melanoma trial (NCT01753089). These preclinical studies illustrate the importance of vaccine selection as well as timing of immunogenic therapies but predicting how they will translate in clinical studies is difficult. Furthermore, choosing among the many possible dosing sequences for testing in a clinical trial is a significant challenge.

Recently, the relationship between mutational burden and neoantigen immunogenicity has been under investigation but is not yet completely understood. Small but consistent data sets have implicated high mutational burden as a biomarker for response to ICI in melanoma [66] and NSCLC [67]. High mutational burden, resulting in increased neoantigen generation, is a putative mechanism of increased responses to checkpoint inhibitors in tumors harboring DNA mismatch repair mutations [68–70]. In 2017, pembrolizumab was approved as the first tissue/site-agnostic drug, given its efficacy in microsatellite instability-high or mismatch repair-deficient unresectable or metastatic solid tumors [71]. In the future, microsatellite instability may become a biomarker for ICI therapy [72], potentially signaling an era when cancer treatment may be dictated not merely by disease type, but also by biomarkers.

Multiple trials in patients with advanced tumors are testing the biomarker approach: NCI-MATCH (Molecular Analysis for Therapy Choice, NCT02465060), IMPACT 2 (Initiative for Molecular Profiling and Advanced Cancer Therapy, NCT 02152254), and NCI-MPACT (Molecular Profiling-Based Assignment of Cancer Therapy, NCT 01827384). Unfortunately, using biomarkers to guide therapy choice is not always predictable or successful, as evidenced by the ineffectiveness of both BRAF and MEK inhibitors in colorectal cancer with BRAF V600 mutations, which have been shown to be effective in advanced melanoma and NSCLC harboring BRAF mutations [73].

7. Five-year view

As discussed above, combinations of ICI with both approved and unapproved vaccines have shown promise as a means to facilitate the immune system and increase efficacy without substantially increasing toxicity in early-phase studies [36–39,41]. Future efforts should aim to include agents that maximize facilitation of the immune system. For example, M7824 is a bifunctional fusion protein that inhibits PD-L1, sequesters transforming growth factor beta (TGF-β), and mediates antibody-dependent cell-mediated cytotoxicity [74]. Preclinical studies have shown that, like PD-L1, TGF-β expressed on cancer cells has immunosuppressive effects in the TME [75,76]. Interesting data by Powles et al. suggest that TGF-β may alter the TME and restrict T cell infiltration [77]. PD-L1 and TGF-β pathways function in an independent and complementary fashion, making combined blockade a promising therapeutic strategy [78,79]. Totally, 19 patients with advanced solid tumors were recently treated in a phase I study of M7824. This concept is being further explored in a clinical trial combining PROSTVAC, M7824, and CV301, a viral vector-based vaccine targeting carcinoembryonic antigen and mucin 1 in men with biochemically recurrent prostate cancer (NCT03315871).

Several other classes of agents under development show promise for combination with vaccines and ICIs. For example, indoleamine 2,3-dioxygenase-1, an enzyme overexpressed in many solid tumors, catalyzes the conversion of tryptophan to N-formyl-kynurenine and enables immune escape [80,81]. Inhibiting this enzyme has shown clinical activity in combination with ICI [82,83]. Engineered cytokines (immunokines) are capable of expanding and activating NK cells and T cells [84]. Preliminary data from PIVOT-02 (NCT02983045) presented at the 2017 Annual Meeting of the Society for Immunotherapy of Cancer highlighted the clinical activity of nivolumab in combination with NKTR-214, a pegylated IL-2 molecule that preferentially binds the IL-2 receptor β-chain. NKTR-214 expands CD8 + T cells and NK cells. These preliminary data highlighted an up to 450-fold increase in tumor-infiltrating CD8 + T cells from baseline, an acceptable safety profile and promising response rates in several malignancies [85]. Other promising agents include immune agonist monoclonal antibodies targeting GITR, 41BB, and OX40.

Key issues.

• Two therapeutic cancer vaccines and six ICIs are currently approved by the U.S. Food and Drug Administration.

• To be successful, a therapeutic cancer vaccine must induce a functional immune response at sites distant from the vaccination site.

• The goal of combining vaccines and ICIs is to modulate the immunosuppressive tumor microenvironment and harness an improved immune response to reduce tumor burden.

• Combination immunotherapy approaches that prime the immune system, expand the immune response, and then facilitate immune function within the TME offer a means to improve the clinical activity of ICIs.

• Most studies of cancer vaccine monotherapies have had negative results, while studies combining vaccines and ICIs have had mixed or moderately positive results.

• Various cancer vaccines have demonstrated the ability to prime the immune system and induce immune responses specific to antigens contained within the vaccine and antigens that are not, a phenomenon known as antigen spreading or antigen cascade.

• Neoantigens are highly expressed in tumor, while tumor-associated antigens are also expressed to some degree in normal tissue. Neoantigen targeting has emerged as a promising strategy for vaccine development.

• Selection of cancer vaccine targets, vaccine adjuvants, and timing of administration of vaccines and ICIs is important for clinical efficacy.

• Recently, tumors with high mutational burden have been shown to have higher response rates to ICIs. The relationship between mutational burden and neoantigen immunogenicity is under investigation.