Clinical Trial on the Safety and Tolerability of Personalized Cancer Vaccines Using Human Platelet Lysate-Induced Antigen-Presenting Cells

Simple Summary In this study, we developed human platelet lysate-induced antigen-presenting cells (HPL-APCs) from peripheral monocytes with a high potency of presentation ability. This study aimed to verify the safety, tolerability, and immunoinducibility of HPL-APCs loaded with cancer candidate antigens as a Phase I study. As a result of the interim analysis, safety and tolerability were confirmed in three enrolled patients, and the immune response to cancer antigen candidate peptides predicted in silico was confirmed in two completed cases. This clinical study is the first to verify the feasibility and immunoinducibility of a personalized cancer vaccine using HPL-APCs that would be expected to demonstrate further antitumor activity through optimized combination therapies. Abstract Research and development of personalized cancer vaccines as precision medicine are ongoing. We predicted human leukocyte antigen (HLA)-compatible cancer antigen candidate peptides based on patient-specific cancer genomic profiles and performed a Phase I clinical trial for the safety and tolerability of cancer vaccines with human platelet lysate-induced antigen-presenting cells (HPL-APCs) from peripheral monocytes. Among the five enrolled patients, two patients completed six doses per course (2–3 × 107 cells per dose), and an interim analysis was performed based on the immune response. An immune response was detected by enzyme-linked immunosorbent spot (ELISpot) assays to HLA-A*33:03-matched KRASWT, HLA-DRB1*09:01-compliant KRASWT or G12D, or HLA-A*31:01-matched SMAD4WT, and HLA-DRB1*04:01-matched SMAD4G365D peptides in two completed cases, respectively. Moreover, SMAD4WT-specific CD8+ effector memory T cells were amplified. However, an attenuation of the acquired immune response was observed 6 months after one course of cancer vaccination as the disease progressed. This study confirmed the safety and tolerability of HPL-APCs in advanced and recurrent cancers refractory to standard therapy and is the first clinical report to demonstrate the immunoinducibility of personalized cancer vaccines using HPL-APCs. Phase II clinical trials to determine immune responses with optimized adjuvant drugs and continued administration are expected to demonstrate efficacy.


Introduction
Advances in next-generation sequencing technology make it possible to treat cancer as precision medicine based on individual cancer genome information. However, only a limited number of patients can benefit from individualized therapy [1], and cancer vaccines based on patient-specific cancer genome information are expected as novel cancer treatments [2].
Cancer vaccines targeting either tumor-associated antigens or tumor-specific antigens, including neoantigens generated by genetic mutations of cancer cells, are being developed [3]. Among cancer vaccine platforms including RNA, DNA, peptides, and dendritic cells (DCs) [4], DC vaccines have been reported to have strong immunoinductibility [5,6]. Immune memory is important for the sustained efficacy of cancer vaccines [7], and the function of DCs that play a central role in immune acquisition is attracting attention [8].
Clinical trials of neoantigen-pulsed DC vaccines using in silico analysis based on cancer genomic information in patients with advanced lung cancer have been conducted [9], and the immune response to neoantigens targeting compatible peptides on HLA class I or II has been confirmed. On the contrary, the immune memory induced by DC vaccination must be comprehensively investigated.
In this study, we verified the safety and tolerability of HPL-APCs loaded with personalized cancer candidate antigens based on cancer genomic profiles and immune memory in a Phase I trial in advanced and recurrent cancers.

Prediction of Cancer Antigen Candidate Peptides
Cancer antigen candidate peptides based on the binding affinities of 8-to 14-mer to HLA-A or 12-to 18-mer to HLA-DPB1/DRB1 were examined using the IEDB analysis resource NetMHCpan (ver. 4.1) [17]. The peptides that predicted the binding affinity the most strongly were selected for the HPL-APC vaccine according to the following priority order: (1) mutated peptides with IC 50 < 500 nM and (2) wild-type peptides with IC 50 < 500 nM.

HPL-APC Vaccination
The release criteria of the HPL-APC vaccine were set at CD86 + HLA-DR + , >70% viability, negative for bacterial and fungal infections, < 1 EU/mL of endotoxin, and negative for mycoplasma. Cryopreserved HPL-APCs were thawed during each session and suspended in 1 mL of saline. The HPL-APC vaccine was intradermally and bilaterally administered near the axillary region and groin.

Evaluation of Safety and Tolerability
The evaluation of safety included the following: (1) any systemic reactions at an early phase within 48 h of the intradermal injection of the HLA-APC vaccine (presence of high fever, hypotension, shock tachycardia, bradycardia, breathing difficulties, or skin rash) and (2) local skin reactions at the injection sites; other symptoms, including nausea, vomiting, diarrhea, appetite loss, mucosal ulcer, or central nervous system disorders; and laboratory test dysfunctions: anemia, reduced white blood cells, reduced platelet count, or elevated kidney and liver serum levels during or after the course of vaccination as described previously [10].

Enzyme-Linked Immunosorbent Spot (ELISpot) Assays
To investigate the antigen-specific IFN-γ production, ELISpot assays were performed using the human IFN-γ ELISpot PLUS kit (Mabtech AB, Nacka Strand, Sweden). Moreover, 1 × 10 6 PBMCs were incubated in 96 wells with AIM-V medium with 10% fetal bovine serum (FBS) and 10 µM peptides. After 16-20 h of incubation, IFN-γ producing cells were detected according to the manufacturer's protocol. The spots were calculated using an automatic ELISpot reader (AID ELISpot Reader Classic ELR 07; Autoimmun Diagnostika GmbH, Strassberg, Germany). The counted numbers with cancer candidate peptides were compared with the HLA-matched HIV peptide (Table S2) as a negative control. The immune response to cancer antigen candidate peptides was defined according to the following criteria: (1) at least 15 spots per 1 × 10 6 PBMCs and (2) at least 1.5-fold more presence of spots than negative control peptide spots [18].

Detection of Antigen-Specific Memory T Cell Subsets
PBMCs were incubated with an AIM-V medium with 10% FBS and 10 µM peptides. HLA-matched HIV peptide was used as the negative control. After 16-20 h incubation, cells were washed and treated with FcR Blocking Reagent (Miltenyi Biotec) and stained with antibodies as shown in Table S5 at 4 • C for 60 min. To detect antigen-specific memory T cells, the activation of CD137 on CD8 + T cells was analyzed [21]. Further, naïve cells as CD45RO − CCR7 + , central memory cells as CD45RO + CCR7 + , effector memory cells as CD45RO + CCR7 − , and effector cells as CD45RO − CCR7 − were identified for memory T cell subsets [22]. A male patient in his 50s was diagnosed with pancreatic cancer (stage IV, KRAS G12D gene mutation). Apheresis was performed without adverse events such as vasovagal reflexes. A course of HPL-APC vaccination for KRAS combined with gemcitabine + proteinbound paclitaxel chemotherapy was completed in six doses, with a total cell count of 14.0 × 10 7 cells (mean viable cell rate, 91.9%). No allergic reaction with delayed-type hypersensitivity (DTH) lower than 5 mm to peptides was noted, and a grade 2 fever reaction and redness of grade 2 at the vaccine-injected sites were observed; however, no other adverse reactions or adverse events were reported. No adverse events were found during the observation period, and follow-up continued. Diagnostic imaging after one course revealed a stable disease (SD); however, 4 months after one course, imaging of the metastasized lung showed a progressive disease (PD), and chemotherapy with 5FU + irinotecan hydrochloride hydrate was performed.

Case 2 (Patient 3)
A female patient in her 40s was diagnosed with cervical cancer (adenocarcinoma, stage IV, KRAS G12D and SMAD4 G365D gene mutations). Apheresis was performed with no adverse events. One course of the HPL-APC vaccine against KRAS and SMAD4 in combination with oral cyclophosphamide was completed, with a total cell count of 17.7 × 10 7 cells (mean viable cell rate, 93.9%) without any quality problems. No allergic reactions with DTH reaction of peptides lower than 5 mm were noted. Grade 2 fever reactions, chills, arthralgia, and redness of grade 2 in the vaccination area were observed but were well tolerated. At the second dose, vaginal bleeding due to primary disease infiltration was observed; however, no other vaccine-related adverse reactions or adverse events occurred. Diagnostic imaging after one course was SD; however, at 4 months after one course, a colostomy was performed because of cancer progression.

Case 3 (Patient 5)
A female patient in her 70s was diagnosed with pancreatic cancer (postoperative recurrence stage IV, KRAS G12D and TP53 E258K gene mutations). Apheresis had already been performed in the WT1 clinical study to obtain materials. During 5FU + irinotecan hydrochloride hydrate salvage therapy, she was administered the HPL-APC vaccine against KRAS and TP53. Neither allergic reaction to DTH nor febrile reactions occurred; however, no adverse reactions or adverse events other than redness of grade 1 were found at the vaccination site. Because the cancer progressed to aortic lymph node metastasis, multiple liver metastases, bile duct obstruction, and portal vein invasion, only bile duct stents and conservative treatment including celiac nerve block were administered without anticancer drugs. After four vaccine lots were administered, she died from the cancer's further progression.

Preparation of Personalized HPL-APC Vaccine with Cancer Antigen Candidate Peptides
By targeting gene products with pathogenic mutations from cancer genomic information, we predicted the affinity of cancer antigen candidate peptides for HLA typing in each patient. Peptides exhibiting an affinity of IC50 < 500 nM were selected for HLA class I and HLA class II (Tables 1 and 2). In patients 2 and 3, killer peptides for the activation of CD8 + T cells with a high affinity for HLA class I of IC 50 < 50 nM were identified (KRAS WT -HLA-A*33:03, IC 50 = 11 nM; SMAD4 WT -HLA-A*31:01, IC 50 = 6 nM). Three enrolled patients had HLA-DRB1*09:01 and were identified to have helper peptides for the activation of CD4 + T cells with a moderate affinity for KRAS G12D of IC 50 < 500 nM (IC 50 = 181 nM, KRAS G12D -HLA-DRB1*09:01). The quality of the HPL-APC vaccine harboring candidate peptides targeting personalized cancer antigens is verified in Table 3. HPL-APCs showed high yields (>23%), viability (>94%), and purity (>91%) in three patients. Phenotypic analysis using a flow cytometer revealed HLA-ABC and HLA-DR expression with costimulatory molecules CD40, CD80, and CD86. These results confirmed the validated quality of HPL-APCs containing cancer antigen-specific peptides for use in clinical research.

Detection of Immune Responses to Personalized Cancer Antigen Candidate Peptides
The interim analysis of immune responses was performed on three enrolled participants, including two who completed six doses per course. Immune responses were monitored by ELISpot assays using personalized cancer antigen candidate peptides (Figure 1). The immune response was evaluated according to the criteria for ELISpot assays [18]. In patient 2, an immune response to HLA-A*33:03-matched KRAS WT was detected after six doses of the HPL-APC vaccine (Figure 1b). Immune responses to KRAS WT peptide compatible with HLA-DRB1*07:01 and KRAS G12D to HLA-DRB1*09:01 were detected ( Figure 1c). Furthermore, an immune response was detected using a KRAS WT -HLA-DRB1*09:01 that had a sequence corresponding to KRAS G12D -HLA-DRB1*09:01. These immune responses attenuated 6 months after completion of HPL-APC vaccination. In patient 3 (Figure 1d-f), an immune response to HLA-A*31:01-matched SMAD4 WT peptide was strongly detected after three and six doses of the HPL-APC vaccine (Figure 1e). Positive reactions to SMAD4 G365D peptide compatible with HLA-DRB1*04:01 and those of KRAS G12D or WT to HLA-DRB1*09:01 were detected (Figure 1f). Six months after one course of the HPL-APC vaccine, the immune response was also diminished. An interim analysis of patient 5 after three doses of the HPL-APC vaccine (Figure 1g

Detection of Immune Responses to Personalized Cancer Antigen Candidate Peptides
The interim analysis of immune responses was performed on three enrolled participants, including two who completed six doses per course. Immune responses were monitored by ELISpot assays using personalized cancer antigen candidate peptides (Figure 1). The immune response was evaluated according to the criteria for ELISpot assays [18]. In patient 2, an immune response to HLA-A*33:03-matched KRAS WT was detected after six doses of the HPL-APC vaccine (Figure 1b). Immune responses to KRAS WT peptide compatible with HLA-DRB1*07:01 and KRAS G12D to HLA-DRB1*09:01 were detected ( Figure  1c). Furthermore, an immune response was detected using a KRAS WT -HLA-DRB1*09:01 that had a sequence corresponding to KRAS G12D -HLA-DRB1*09:01. These immune responses attenuated 6 months after completion of HPL-APC vaccination. In patient 3 (Figure 1d-f), an immune response to HLA-A*31:01-matched SMAD4 WT peptide was strongly detected after three and six doses of the HPL-APC vaccine (Figure 1e). Positive reactions to SMAD4 G365D peptide compatible with HLA-DRB1*04:01 and those of KRAS G12D or WT to HLA-DRB1*09:01 were detected (Figure 1f). Six months after one course of the HPL-APC vaccine, the immune response was also diminished. An interim analysis of patient 5 after three doses of the HPL-APC vaccine (Figure 1g

Discussion
This study confirmed the safety and tolerability of the HPL-APC vaccine in two patients with pancreatic cancer (patients 2 and 5) and one with cervical cancer (patient 3) and clarified the immunoinductibility to personalized cancer antigen candidate peptides using ELISpot assays. HLA class I-compatible killer peptides or HLA class II-compliant helper peptides targeting KRAS, SMAD4, or TP53 were predicted from cancer genomic profiles (Table 2), which are pancreatic cancer driver genes frequently identified as pathogenic mutations [23]. Immunoinductibility to KRAS WT or G12D peptides for HLA class II was confirmed in patient 2 after six sessions, and that to SMAD4 WT peptide for HLA class I and SMAD4 G365D peptide for HLA class II in patient 3 were detected after three doses

Discussion
This study confirmed the safety and tolerability of the HPL-APC vaccine in two patients with pancreatic cancer (patients 2 and 5) and one with cervical cancer (patient 3) and clarified the immunoinductibility to personalized cancer antigen candidate peptides using ELISpot assays. HLA class I-compatible killer peptides or HLA class II-compliant helper peptides targeting KRAS, SMAD4, or TP53 were predicted from cancer genomic profiles (Table 2), which are pancreatic cancer driver genes frequently identified as pathogenic mutations [23]. Immunoinductibility to KRAS WT or G12D peptides for HLA class II was confirmed in patient 2 after six sessions, and that to SMAD4 WT peptide for HLA class I and SMAD4 G365D peptide for HLA class II in patient 3 were detected after three doses (Figure 1). Although these peptides for the induction of antigen-specific immune response to CD8 + and CD4 + T cells in PBMCs after the HPL-APC vaccine were suggested, future analysis is needed to clarify the responding cells from IFN-γ production using flow cytometry. The activation of cancer antigen-specific CD8 + T and CD4 + T cells is a critical issue for the efficacy of cancer vaccines [7]. KRAS is the most frequently mutated isoform present in 22% of all tumors [24], and KARS G12D is the highest frequency of approximately 35% among them [25]. Immunoinductibility could not be observed for KRAS G12D -HLA-A*11:01 (VVGADGVGK, 9-mer peptide) in patient 3 (Figure 1d). However, as the TCR repertoire responding to VVVGADGGGK (10-mer peptide) to HLA-A*11:01 binding KRAS G12D peptide was reported [26], the immunoinductibility of HPL-APCs to these peptides must be verified in the future with an additional number of cases. On the contrary, KRAS G12D -HLA-DRB1*09:01 helper peptide common to all three cases was chosen (TEYKLVVVGADGVGK, 15-mer peptide in Table 2). Although a TCR repertoire for HLA-DRB1*08:01 binding KRAS G12D was reported [27], immunoinduction against KRAS G12D -HLA-DRB1*09:01 helper peptide was first identified; however, the TCR repertoire has not yet been analyzed. Interestingly, an immune response was also detected when using a KRAS WT -HLA-DRB1*09:01 that had a sequence corresponding to KRAS G12D -HLA-DRB1*09:01 (Figure 1c,f). The TCR repertoire analysis induced by the KRAS G12D -HLA-DRB1*09:01 helper peptide is expected to reveal specificity and further verify the differences in the antitumor response.
DC vaccines can enhance effector and memory T cell responses, which are important for antitumor immunity and have been used in clinical trials of cancer immunotherapy [28,29]. A Phase I clinical trial of a DC vaccine pulsed with Wilms' tumor 1 (WT1)specific MHC class I/II-restricted epitope for pancreatic cancer in combination with chemotherapy was reported. An association was found between WT1/HLA-A*24:02specific memory cells in CD8 + T cells and overall survival (OS) [12]. WT1-specific central and effector memory CD8 + T cells were observed in cancerous pleural effusion in a patient with long-survival pancreatic cancer after WT1-pulsed DC vaccination [16]. In Phase II trials to evaluate the efficacy of personalized cancer vaccines using HPL-APCs, the detection of antigen-specific memory T cells could be a biomarker of therapeutic prognosis. In patients 2 and 3, CD8 + T EM increased after three doses of the HPL-APC vaccine ( Figure S2), and in patient 3, an increase in antigen-specific CD8 + T EM was observed with SMAD4 WT -HLA-A*31:01 peptide in a small number of cells due to limited clinical specimens ( Figure 2b). Because T EM has strong cytotoxicity with high cytokine production capacity [30], it was consistent with the increase in IFN-γ-producing cells observed with SMAD4 WT HLA-A*31:01 peptide (Figure 1e). These results confirmed the ability of HPL-APCs to enhance antigen-specific CD8 + memory T cells. On the contrary, the detection of CD8 + memory T cells to KRAS WT -HLA-A*33:03 and that of CD4 + memory T cells responding to KRAS WT or G12D or SMAD4 G365D helper peptides remain to be elucidated. A method for highly efficient detection of neoantigen-responsive T cells and identification of these TCRs has been reported for personalized cancer immunotherapy [31]. Cancer antigen-specific memory T cells induced by the HPL-APC vaccine using a method that can be easily detected from the peripheral blood via TCRs as an indicator must be identified.
In patients 2 and 3, who completed the HPL-APC vaccination, the acquired immune response was attenuated 6 months after HPL-APC vaccination (Figure 1b,c,e,f), and in patient 2, a significant decrease in T EM was observed ( Figure S2). In patients with long-term survival receiving continuous WT1-DC vaccine, acquired WT1-CTLs were maintained [32], and booster effects of additional vaccines were observed [33]. As DC vaccines have been expected to have antitumor effects by priming and boosting [34], the immune memory acquired by the HPL-APC vaccine must be also preserved. Therefore, activating the immune response with additional administrations of the HPL-APC vaccine is expected. As the life span of DC in the lymph nodes is limited to a few days [35,36], the persistence of the HPL-APC vaccine also needs to be monitored in patients to ensure long-lasting efficacy.
On the contrary, the immunosuppressive factor M-MDSCs increased, and no immune response was observed in patient 5 (Figure 1g,h). The combination of DC vaccines with chemotherapy is important to create a state favoring expansion of antitumor effector cells, and inhibition, depletion, or both, of Tregs and MDSCs [37]. In a future Phase II study to verify the efficacy of the HPL-APC vaccine, the antitumor immune reactions in combination with anticancer drugs that eliminate or abolish the function of M-MDSCs and immune checkpoint inhibitors that restore exhausted T cells must be optimized [38].

Conclusions
This is the first Phase I clinical trial that demonstrates the safety, tolerability, and immunoinductibility of the HPL-APC vaccine with personalized cancer antigen candidate peptides. The HPL-APC vaccine can enhance antigen-specific CD8 + memory T cells. The maintenance of immune memory might require continued HPL-APC vaccination and improvement of the immunosuppressive environment. The feasibility of personalized cancer vaccines using HPL-APCs has been clarified, and a future Phase II trial with optimized dosing and combination therapies would be expected to provide progress in antitumor response.

Institutional Review Board Statement:
The HPL-APC vaccination study in patients with cancer (approval No. PC009; 18 November 2021) was approved by the Kanazawa Medical University Certificated Committee for Regenerative Medicine (Class III technologies) (approval No. of the committee NB4150006) and was registered as jRCTc040210109 in the Japan Registry of Clinical Trials. All investigations were performed according to the Declaration of Helsinki. An accompanying study of HPL-APC vaccination therapy to monitor the immune response was conducted after receiving appropriate informed consent. This study was approved by the Ethical Committee of Kanazawa Medical University (approval No. G156; 8 June 2020).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available in the article and Supplementary Materials.

Conflicts of Interest:
S.S., T.K. and M.T. are patent inventors for a method for manufacturing human platelet lysate-inducible interferon (IFN)-DCs (PCT/JP/2021/040505). S.S., T.K. and M.T. evaluated the DCO-K serum-free medium (Nissui Pharmaceutical Co., Ltd.) for the HPL-APC culture. Kanazawa Medical University concludes a collaborative investigation contract with Nissui Pharmaceutical Co., Ltd. The funders had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.