Study design
This is a single‐arm, Phase 1b, open‐label clinical trial for patients with MIBC (cT2‐4), with/out pelvic lymphadenopathy (cN0‐1) and no evidence of distant metastases (cM0) who are cisplatin ineligible or who refused cisplatin prior to radical cystectomy. The trial is sponsored by CG Oncology and has been approved by the Moffitt Cancer Center (MCC) Institutional Review Boards (IRB) (MCC20575) and registered at ClinicalTrials.gov (NCT04610671). The primary objective is to test the safety of this drug combination, assessed per the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0. Dose‐limiting toxicities (DLTs) were assessed in 6 DLT‐evaluable patients in a safety lead‐in phase. The combination was declared tolerable as the incidence of DLTs was <33%, and the study expanded to 30 total patients. Secondarily, clinical measures of efficacy related to the drug combination will be assessed. An amendment in December 2022 reduced the sample size to 21 patients as both the sponsor and investigator agreed that safety of the regimen and early efficacy signal had been established, allowing expedited planning for more definitive studies.
Patients
Patients were cisplatin-ineligible as per the following definition: glomerular filtration rate < 60mL/min, chronic heart failure New York Heart Association class III or higher, peripheral neuropathy grade 2 or higher, Eastern Cooperative Oncology Group score ≥2 or impaired hearing. Patient’s refusal of traditional chemotherapy was also included. Other inclusion/exclusion criteria are outlined in the protocol (see Supplemental Information). During screening, patients were evaluated by a surgeon to determine eligibility for RC-PLND, as well as undergoing complete tumor staging using CT thorax/abdomen/pelvis or non‐contrast CT/MRI (in case of poor renal function). Screening cystoscopy was performed at baseline to evaluate location of the tumor, and residual tumor burden. All patients provided informed consent.
Inclusion Criteria
- Patients must have histologically confirmed MIBC (cT2‐T4a, N0‐N1, M0 per American Joint Commission on Cancer [AJCC]) pure or mixed histology urothelial carcinoma. Clinical node‐ positive (N1) patients are eligible provided the lymph nodes are within the planned surgical dissection template.
- The initial TURBT that showed muscularis propria invasion should be within 90 days prior to beginning study therapy. Patients must have sufficient baseline tumor tissue from either initial or repeat TURBTs. The local site pathologist will be asked to estimate and record the rough approximate percentage of viable tumor in the TURBT sample (initial or repeat TURBT with highest tumor content) to document at least 20% viable tumor content prior to registration. This is to ensure adequate tissue is available to perform tumor infiltrating CD8+ T‐cell assessment.
- Patients must be ineligible for cisplatin‐based chemotherapy due to any of the following:
- Creatinine clearance (CrCl) < 60 mL/min (with ECOG Performance Status 0‐1)
- Hearing impaired ≥ Grade 2 by CTCAE criteria
- Neuropathy ≥ Grade 2 by CTCAE criteria
- Heart failure NYHA ≥ III
- ECOG ≥ 2
- Refusing to undergo cisplatin chemotherapy
- Patients must be medically fit for TURBT and radical cystectomy.
- Age ≥ 18 years.
- Ability to understand and willingness to sign IRB‐approved informed consent.
- Willing to provide tumor tissue, blood, and urine samples for research.
- Adequate organ function as measured by the following criteria, obtained ≤ 4 weeks prior to registration:
- Absolute Neutrophil Count (ANC) ≥ 1,000/mm3 (stable off growth factor within 4 weeks of first study drug administration)
- Platelets ≥ 100,000/mm3
- Hemoglobin ≥ 8 g/dL
- Serum Creatinine Clearance ≥ 20 mL/min using the Cockcroft‐Gault formula
- ALT and AST ≤ 2.5x ULN
- Total Bilirubin ≤ 1.5x ULN (in the absence of diagnosed Gilbert’s disease)
Exclusion Criteria
- Patients who are pregnant or breastfeeding, since the effects of nivolumab and CG0070 on the fetus or breastfeeding child are unknown. All sexually active females of childbearing potential (not surgically sterilized and between menarche and 1 year post menopause) must have a blood test to rule out pregnancy within 4 weeks prior to registration.
- Patient with local symptoms from bladder cancer, (e.g. gross hematuria, dysuria, etc.) who are deemed to be unable to complete the treatment protocol.
- Patients with active or prior documented autoimmune disease within the past 2 years prior to Screening or other immunosuppressive agent within 14 days of study treatment.
NOTE: Patients with well controlled type 1 diabetes mellitus, vitiligo, Graves disease, Hashimoto’s disease, eczema, lichen simplex chronicus, or psoriasis not requiring systemic treatment (within the past 2 years prior to Screening) are not excluded.
- Patients who have concurrent upper urinary tract (i.e. ureter, renal pelvis) invasive urothelial carcinoma. Patients with history of non‐invasive (Ta, T1, Tis) upper tract urothelial carcinoma that has been definitively treated with at least one post‐treatment disease assessment (i.e. cytology, biopsy, imaging) that demonstrates no evidence of residual disease are eligible.
- Patients who have another malignancy that could interfere with the evaluation of safety or efficacy of the study drugs. Patients with a prior malignancy will be allowed without Principal Investigator approval in the following circumstances:
- Not currently active and diagnosed at least 3 years prior to the date of registration.
- Non‐invasive diseases such as low risk cervical cancer or any cancer in situ.
- Localized (early stage) cancer treated with curative intent (without evidence of recurrence and intent for further therapy), and in which no chemotherapy was indicated (e.g. low/intermediate risk prostate cancer, etc.). Patients with other malignancies not meeting these criteria must be discussed prior to registration.
- Patients who have received any prior immune checkpoint inhibitor (i.e. anti‐KIR, anti‐PD‐1, anti‐PD‐L1, anti‐CTLA4 or other).
- Patients who have undergone major surgery (e.g. intra‐thoracic, intra‐abdominal or intra‐pelvic), open biopsy or significant traumatic injury or specific anti‐cancer treatment ≤ 4 weeks prior to starting study drug, or patients who have had placement of vascular access device ≤ 1 week prior to starting study drug, or who have not recovered from side effects of such procedure or injury.
- Patients who have clinically significant cardiac diseases deemed not fit for radical cystectomy, including any of the following:
- History or presence of serious uncontrolled ventricular arrhythmias.
- Clinically significant resting bradycardia.
- Any of the following within 3 months prior to starting study drug: severe/unstable angina, Congestive Heart Failure (CHF), Cerebrovascular Accident (CVA), Transient Ischemic Attack (TIA).
- Uncontrolled hypertension defined by a SBP ≥ 180 mm Hg and/or DBP ≥ 100 mm Hg, with or without anti‐hypertensive medication(s).
- Patients who have history of chronic active liver disease or evidence of acute or chronic Hepatitis B Virus (HBV) or Hepatitis C (HCV).
- Patients who have known diagnosis of human immunodeficiency virus (HIV) infection. Testing is not required in absence of clinical suspicion.
- Patients who have known diagnosis of any condition (e.g. post‐hematopoietic or solid organ transplant, pneumonitis, inflammatory bowel disease, etc.) that requires chronic immunosuppressive therapy which cannot be stopped for the duration of the clinical trial. Usage of non‐steroidal anti‐inflammatory medications (NSAIDS) for the treatment of osteoarthritis and uric acid synthesis inhibitors for the treatment of gout are permitted.
- Patients with any serious and/or uncontrolled concurrent medical conditions (e.g. active or uncontrolled infection, uncontrolled diabetes) or psychiatric illness that could, in the investigator’s opinion, cause unacceptable safety risks or potentially interfere with the completion of the treatment according to the protocol.
- Patients who have used any live viral vaccine for prevention of infectious diseases within 4 weeks prior to study drug(s). Examples of live vaccines include, but are not limited to, the following: measles, mumps, rubella, varicella/zoster (chicken pox), yellow fever, rabies, BCG, and typhoid vaccine. Seasonal influenza vaccines for injection are generally killed virus vaccines and are allowed; however, intranasal influenza vaccines (e.g., FluMist®) are live attenuated vaccines and are not allowed.
- Patients unwilling or unable to comply with the protocol.
- Patients with a known allergy to any of the study medications, their analogues, or excipients in the various formulations of any agent.
- Patients who participate in any other therapeutic clinical trials, including those with other investigational agents not included in this trial throughout the duration of this study.
- Use of excluded antiviral medication that cannot be suspended at least 14 days prior and for 14 days after the administration of any CG0070 treatment throughout the duration of the trial.
Treatment
Intravesical cretostimogene grenadenorepvec (CG0070) was administered following a sequence of bladder washes with normal saline and n‐dodecyl‐B‐D‐maltoside (DDM), within 90 days from the biopsy demonstrating MIBC and in the absence hematuria/traumatic catheterization. Cretostimgene (1x1012 viral particles, vp) was administered weekly x 6 and nivolumab (480mg IV) at weeks 2 and 6. A post‐treatment TURBT was strongly recommended within 3-4 weeks of treatment completion. RC- PLND occurred 2 weeks following post-treatment TURBT. All subjects were followed clinically and radiographically per standard-of-care following radical cystectomy. Pre-, on-, and post-treatment blood, urine and tissue samples were collected for correlative biological analyses per an IRB-approved laboratory protocol (MCC 22288).
GM-CSF expression
To measure the amount of GM-CSF levels following cretostimogene treatment, primary urine samples from eligible patients were collected at different time points including at baseline, week 2 and week 6 pre-treatment. Urine specimens were spun down at 10,000xg and supernatants were saved for ELISA assays. For the ELISA experiments, QuantikineTM HS human GM-CSF kit was used (R&DSystems Inc/Biotechne, Cat # HSGMO). The assay was performed in triplicate and according to manufacturer’s recommendations.
Olink analysis on urine samples
Urine samples were analyzed for the presence of inflammation and immuno-oncology related proteins using the Olink Target 96 Inflammation, Immuno-Oncology, and Oncology Assays (Olink). In short, these measurements are based on the Proximity Extension Assay (PEA) technology which enables high-throughput multiplex immunoassays of proteins using 1 µl of urine. Normalized Protein eXpression (NPX) median-based intensity normalized data provided by OlinkⓇ was used for data analysis. Samples not passing quality control were removed. Assays were excluded if less than 20% of the samples were above limits of detection (LOD). LOD for assays ranged between -0.53 and 4.98.
Fluorescent multiplex immunohistochemistry (IHC) panel
Formalin-fixed and paraffin-embedded (FFPE) tissue samples were immune-stained using the AKOAYA Biosciences OPAL TM 7-Color Automation IHC kit (Waltham, MA) on the BOND RX auto-Stainer (Leica Biosystems, Vista, CA). The OPAL 7-color kit uses tyramide signal amplification (TSA)-conjugated to individual fluorophores to detect various targets within the multiplex assay. Sections were baked at 65oC for one hour then transferred to the BOND RX (Leica Biosystems). All subsequent steps (e.g., deparaffinization, antigen retrieval) were performed using an automated OPAL IHC procedure (AKOYA). OPAL staining of each antigen occurred as follows: heat induced epitope retrieval (HIER) was achieved with Citrate pH 6.0 buffer for 20 min at 95°C before the slides were blocked with AKOYA blocking buffer for 10 min. Then, slides were incubated with primary antibody, CD138 (Abcam, EPR6454, 1:700, dye 620) at RT for 60 min followed by OPAL HRP polymer and one of the OPAL fluorophores during the final TSA step. Individual antibody complexes are stripped after each round of antigen detection. This was repeated five more times using the following antibodies; CD8 (DAKO, C8/144B, HIER-EDTA pH 9.0, 1:50, dye520), CD4 (CM, EP204, HIER- EDTA pH 9.0, 1:100, dye570), CD20 (DAKO, L26, HIER-EDTA pH 9.0, 1:300, dye 480), CD3 (DAKO, Rb poly, HIER- EDTA pH 9.0, 1:100, dye690), and PCK (DAKO, AE1/AE3, HIER- Citrate pH 6.0, 1:150, dye780). Additional Multiplex IHC panel were develop using the following antibodies: PIGR (Abcam, Rb poly 1:50), IgM (Abcam, IM260 1:50), IgA (Abcam, Rb poly, 1:1000) and IgG (Abcam, EPR4421, 1:300). After the final stripping step, DAPI counterstain was applied to the multiplexed slide and was removed from BOND RX for cover-slipping with ProLong Diamond Antifade Mountant (ThermoFisher Scientific). All slides were imaged with the Vectra®3 Automated Quantitative Pathology Imaging System.
Quantitative image analysis
Multi-layer TIFF images were spectrally unmixed in InForm version 2.6.0 (AKOYA) and imported into HALO (Indica Labs, New Mexico) for quantitative image analysis. Each image was annotated by an experienced image analysis specialist to identify individual TLS objects by manual creation of regions of interest around areas consisting of a dense co-clustering of CD3 and CD20 positive cells with an emphasis on the CD20 cell density. Manual scoring was also performed on each TLS to categorize them into immature and mature structures (Suppl. Fig. 5). Criteria for scoring included overall cell density (DAPI channel), shape (distribution of CD3/CD20 cells into circular or ovular formation), and size (greater than 5000 µm2). Immature TLS scoring consisted of less dense structures with unremarkable and undefined shape, while mature TLS were identified as very dense structures highly populated with CD20+ with well-defined shape and morphology. Automated segmentation of Tumor, Stroma and non- tissue regions were performed on the whole tissue image by training a classier using the Pan-cytokeratin (PanCK) channel as a tumor region marker. The classifier was created and tested on various images in the image set. The tissue is segmented into individual cells using the DAPI marker which stains cell nuclei. For each marker, a positivity threshold within the nucleus or cytoplasm were determined per marker based on published staining patterns and intensity for that specific antibody. After setting a positive fluorescent threshold for each staining marker, the TLS, tumor, and stroma regions for the entire image set was analyzed with the created algorithm. The generated data includes positive cell counts for each fluorescent marker in cytoplasm or nucleus, and percent of cells positive for the marker. Along with the summary output, a per-cell analysis can be exported to provide the marker status, classification, and fluorescent intensities of every individual cell within an image. The following table describe the fluorescent antibodies used for the TLS panel.
Reagent
|
Lot#
|
Product#
|
Opal
|
Clone
|
Dilution
|
Manufacturer
|
AR
|
Target Antibody 1 CD138
|
1009209-1
|
ab128936
|
OPAL-620
|
EPR6454
|
1:700
|
Abcam
|
ER2
|
Target Antibody 2 CD8
|
41527763
|
M7103
|
OPAL-520
|
C8/ 144B
|
1:50
|
Dako
|
ER2
|
Target Antibody 3 CD4
|
186600
|
104R-25
|
OPAL-570
|
EP204
|
1:50
|
Cell Marque
|
ER2
|
Target Antibody 4 CD20
|
20042864
|
M0755
|
OPAL-480
|
L26
|
1:300
|
Dako
|
ER2
|
Target Antibody 5 CD3
|
41495387
|
A0452
|
OPAL-690
|
Rb poly
|
1:100
|
Dako
|
ER2
|
Target Antibody 6 PCK
|
11478870
|
M3515
|
OPAL-780
|
AE1/ AE3
|
1:150
|
Dako
|
ER2
|
The following table describe the fluorescent antibodies used for the Ig panel.
Reagent
|
Lot#
|
Product#
|
Opal
|
Clone
|
Dilution
|
Manufacturer
|
AR
|
Target Antibody 1 PIGR
|
GR3357008-9
|
ab96196
|
OPAL-480
|
Rb poly
|
1:50
|
Abcam
|
ER2
|
Target Antibody 2 CD20
|
20042864
|
M0755
|
OPAL-570
|
L26
|
1:200
|
Dako
|
ER2
|
Target Antibody 3 IgM
|
GR3345045-2
|
ab200541
|
OPAL-620
|
IM260
|
1:50
|
Abcam
|
ER2
|
Target Antibody 4 IgA
|
GR271550-17
|
ab97216
|
OPAL-520
|
Rb poly
|
1:1000
|
Abcam
|
ER1
|
Target Antibody 5 IgG
|
GR3258727-10
|
ab109489
|
OPAL-690
|
EPR4421
|
1:300
|
Abcam
|
ER1
|
Target Antibody 6 PCK
|
11375363
|
M3515
|
OPAL-780
|
AE1/AE3
|
1:150
|
Dako
|
ER1
|
Tissue preparation followed by DNA and RNA extraction
Tissues were fixated by placing them in 10% neutral buffered formalin within 15 minutes of collection and fixed for 24 hours prior to embedding in paraffin using standard methodologies. Sections (4mm) from each formalin-fixed, paraffin embedded (FFPE) tissue was stained with hematoxylin and eosin (H&E) and reviewed by the study Pathologist (JD) to ensure tumor content. Scroll (3-5, 20mm sections) from each FFPE block were used for total RNA extraction using the Qiagen RNeasy FFPE kit (cat#73504, Qiagen, Germantown, MD) in accordance with manufacturer’s instructions. RNA quality was assessed using Qubit Fluorometric Quantification (Invitrogen) and Tapestation 4200 (Agilent).
Tissues were snap frozen within 15 minutes of collection and stored in liquid nitrogen until processing. Sections (4mm) from each frozen sample were stained with hematoxylin and eosin (H&E) and reviewed by the study Pathologist to ensure tumor content. Thirty milligrams of tissue were macrodissected to enrich for tumor and used for total DNA extraction using the Qiagen DNeasy Blood & Tissue Kit (cat#69504, Germantown, MD) in accordance with manufacturer’s instructions. DNA quality was assessed using Qubit F luorometric Quantification (Invitrogen) and Tapestation 4200 (Agilent).
RNA-seq data analysis
The raw RNA-seq reads were first assessed for quality using FastQC. Quality trimming was performed using cutadapt 44 to remove reads with adaptor contaminants and low-quality bases. Read pairs with either end too short (<25bps) were discarded from further analysis. Next, trimmed and filtered reads were aligned to the human transcriptome GRCh37 using STAR.45 Gene expression was quantified as Transcript per Million (TPM) using RSEM.46 Batch effects were corrected with Combat47 and differential expression analysis was performed by limma on the log-transformed batch-corrected TPM values (Suppl. Fig.2 RNAseq). Genes were then ranked based on –log10(p-value)*[sign of log2(fold-change)], such that the most up-regulated genes were at the top and most down-regulated ones were at the bottom. The pre-ranked gene list was used to perform pre-ranked gene set enrichment analysis (GSEA48 version 4.0.2) to assess enrichment of hallmarks, curated gene sets, and gene ontology49 terms in MSigDB.48,50 The resulting normalized enrichment score (NES) and FDR controlled p-values were used to assess transcriptome changes. Additionally, we employed NMF-based methodologies to delineate the molecular subtypes within tumors, utilizing log-transformed and median centralized TPM value. Briefly, a collection matrix (WTCGA) of 354 subtype-specific differentially overexpressed marker genes and their weights to the 5 subtypes (Luminal, Luminal-Infiltrated, Basal-Squamous, Neuronal, Luminal-Papillary) were identified from TCGA advanced urothelial cancers through unsupervised clustering.51 Utilizing the NMF-based subtype classification model, the bladder cancer subtype for each tumor was determined by approximating X~ WTCGA *H, where X represents the tumor gene expression and H represents the subtype weight vector. The classification was carried out using the R package ADAPTS52 employing non-negative least squares deconvolution.
Somatic mutations identification and profiling
Somatic mutations were identified from whole exome sequencing of matched tumor and normal samples following the strategies described in TCGA’s Multi-Center Mutation Calling in Multiple Cancers project (MC3 project).53 Short reads were aligned to human reference genome GRCh37 using the Burrows-Wheeler Aligner (BWA54) and then improved by base quality score recalibration, sequence realignment near indels, and duplicate read removal using the Genome Analysis Tool Kit version 4 (GATK55) and Picard. SAMSTAT56 and Picard were used for quality checking of the aligned BAM files. Somatic mutations were further detected from the recalibrated bam using a combination of Mutect256, SomaticSniper57, MuSe58, FreeBayes 59, Pindel60, and Strelka61 with the default settings. Mutations with reported variant allele frequency > 0.01 in external databases including 1000 Genomes Project62, NHLBI Exome Sequence Project (ESP) and the Exome Aggregation Consortium (ExAC)63 were considered germline inherited variations and removed from further analysis. Only point mutations and indels predicted by at least two algorithms were included for subsequent analysis. Additional annotation were added from COSMIC64, ExAC63 , dbGAP65, and Ensembl66 using ANNOVAR.67 The identified somatic mutations were further analysed and visualized using R packages Maftools.68 Tumor mutation burden (TMB) was calculated as number of non- synonymous mutations for each tumor. Genes under positive selection as putative oncogenic drivers were identified using dNdScv.69 Four genes (TP53, RB1, ARID1A, KMT2D) with global q-value (qglobal_cv) <0.15 were considered significant. Mutational Signatures were identified by evaluating patterns of nucleotide substitutions using an NMF-based method implemented in Maftools. The top 4 signatures that optimally represented the mutation profile were extracted and matched to SBS signature database.70
Neoantigen prediction
Neoantigens were predicted from somatic mutations detected as described above. The 4-digit patient-specific HLA haplotypes are determined in silico using arcasHLA71 (RNAseq), hlahd72 (WES), and T1K73 (both). All nonsynonymous somatic mutations identified above were translated into 8-14-mer peptides flanking the mutant amino acid for MHC class I (MHC-I) and 12-16-mer peptides flanking the mutant amino acid for MHC class II (MHC-II) neoantigens. Only the mutated sequences with TPM ≥1 in tumor RNAseq data were further analyzed to predict MHC-peptide binding affinity against patient-specific HLA type using NetMHCpan74 for MHC-I and NetMHCIIpan75 for MHC-II. Neoantigen burden was the calculated number of 25-mers mutated peptides with any 9-14-mer or 12-16-mer predicted with higher binding compared to their reference counterparts for either MHC-I or MHC-II.
Neopeptide prioritization
We utilized a two-pass approach to prioritize neopeptides. First, the 25-mer mutated peptides were prioritized based on expression and MHC binding prediction analyses to obtain an Additive Score (AS) as previously described.76 The Expression Score (ES) component was determined by using the maximum variant allele frequency in the RNAseq data (VAF) and fragments per kilobase of exon per million mapped fragments (FPKM) obtained from the RNASeq data. Additionally, the MHC Combined Score (MCS) component incorporated the maximum predicted binding of each peptide to the patient’s MHC molecules, along with the Differential Agretopicity Index (DAI) between the variant (var) peptide and its corresponding reference (ref) peptide for both MHC-I (mhc1) and MHC-II (mhc2). The complete formula for the AS prioritization of 25-mers is provided below:
AS=ES+(MCS/1.5).
ES=max_VAF_RNA_percentile+(0.5×max_FPKM_percentile).
MCS=mhc1_score+(0.5×$mhc2_score).
mhc1_score=ic50_mhc1_percentile+(0.5×DAI_mhc1_percentile).
mhc2_score=ic50_mhc2_percentile+(0.5×DAI_mhc2_percentile).
percentile=max percentile for a given peptide within 25-mer (excluding 0/NA).
DAI: var_ic50/ref_ic50.
In the second pass, within the top 25-mer peptides prioritized, the binding of each 9-14-mer or 12-16-mer neopeptides and their reference counterparts to MHC-I and MHC-II were evaluated as strong binding (SB), weak binding (WB), and no binding (NB) based on %Rank score generated by NetMHCpan and NetMHCIIpan. %Rank < 0.5% and %Rank < 2% thresholds were considered for detecting SBs and WBs for MHC-I and %Rank < 2% and %Rank < 10%, for SBs and WBs for MHC-II. %Rank >=2% and %Rank >=10% were considered as NBs for MHC-I and MHC-II, respectively. We then chose the best 9-14-mer neopeptides for MHC-I and 12-16-mer neopeptides for MHC-II using following criteria: 1) when a neopeptide was classified as SBs, the reference form of the neopeptides was required to be either NBs or WBs with log2(IC50 of reference /IC50 of neopeptide) > 1.2; 2) when a neopeptide was classified as WBs, the reference form of the neopeptides was required to be NBs.
ELISpot assays to quantify the systemic reactivity
A detailed schema is shown in Fig. 3d. PBMCs from enrolled patients were collected at baseline, week 2 and week 6 pre-treatment.
Generation of autologous dendritic cells
The plastic adherence method was used to generate monocyte-derived DCs according to published protocol with a some modifications.22 in brief about 6-9 × 106 PBMCs were suspended in AIM-V medium (Fisher Scientific) and seeded in a tissue culture plates of an appropriate size. Cells were then incubated in a humidified 37oC 5% CO2 incubator for 2 hours. The non-adherent cells (CD14- a source of T cells) were collected and cryopreserved in LN2. Adherent cells were washed with AIM-V medium twice with an interval of 60 min, after which DC medium were added. DC medium was made from CorningTM RPMI 1640 (Fisher Scientific) supplemented with 5% heat inactivated human AB serum plasma (BioIVT-Elevating Science), 100 U ml−1 penicillin and 100 μg ml−1 streptomycin, 2 mM L-glutamine (media supplements were from Fisher Scientific), 80 ng/mL GM-CSF and 160 ng/mL IL-4 (cytokines from Peprotech). On day 3, fresh DC complete medium supplemented with GM-CSF and IL-4 was added to DCs. DCs were collected on day 6 for co-culture experiments.
Screening T cells for reactivity
For the ELISpot assay, human IFN-g ELISpot kit containing 96-PVDF-backed microplate already pre-coated with monoclonal antibody specific for human IFN-g (Cat# IL285 from R&D systems Inc) was used. The ELISpot assay was performed according to the manufacture’s recommendations. On the day prior to co-culture experiments, autologous CD14- non-adherent cells (source of T cells) were thawed and seeded in appropriately sized tissue culture plates in complete RPMI-1640 medium supplemented with 5% heat inactivated human AB serum plasma (BioIVT-Elevating Science), 100 U ml−1 penicillin and 100 μg ml−1 streptomycin, 2 mM L-glutamine (Fisher Scientific) and incubated in a 37oC, 5% CO2 incubator for one overnight (about 18 hours). Before co-culture, wells from the IFN-g precoated microplate were filled with 300 mL of AIM-V culture medium and incubated for 20 minutes at room temperature. When cells were ready to be plated, culture media was aspirated from the wells and immediately after this, 100 mL of appropriate cells or controls suspended in AIM-V culture medium were added to the wells. In fbrief, a cell-to-cell ratio of 1:10 stimulated mo-DCs (1×104) and autologous T cells (1×105) was used. These co-cultures were spiked with the corresponding manufactured predicted neoantigen peptide (GenScript) at a concentration of 1.0 mg/mL. As a negative control, co-cultures were spiked with vehicle control (e.g., 1x-PBS). Co-cultures were incubated at 3oC, 5% CO2 for 24 hours. After 24 hours of co-culture, the content of each well was aspirated and washed 4x-times with wash buffer provided by the ELISpot kit. One hundred (100 mL) of diluted detection antibody mixture was added to each well of the microplate and incubated for 2 hours at room temperature on a rocking platform. After this, wells were washed 4x-times with wash buffer. One hundred (100 mL) of diluted Streptavidin-AP concentrate A was added into each well of the microplate and incubated for 1 hour at room temperature. Wells were washed 4x-times with wash buffer as described before. One hundred (100 mL) of BCIP/NBT substrate was added into each well. After 1 hour of incubation at room temperature for 1 hour and protected from light, BCIP/NBT substrate was decanted, and the wells were rinsed with deionized water. The microplate was completely dried at room temperature for 90 minutes. The microplate was scanned using an ImmunoSpot ELISpot plate reader (Mabtech Apex 2).
For the ELISpot assays, four patients were excluded for reasons including: not enough tissue sample was available for RNA/DNA extraction and therefore no predictions of neoantigens were performed; one of the patients did not complete the clinical study. At some time points there was not sufficient PBMCs to generate autologous mo-DCs and T cells.
Statistics
Adverse events were continuously monitored through 100 days of discontinuation of experimental drug dosing. Acceptable safety was prospectively defined as a dose-limiting toxicity rate of 33% or less, using a Pocock-type stopping boundary with continuous monitoring.77 Dose-limiting toxicity was defined as grade ≥3 toxicity attributed to the treatment agents, detailed in the Supplemental Information. Pearson’s correlation test was used to determine R2. P values were generated using either Wilcoxon or Pearson’s test. Reported P values were two-sided with a significance of 0.05, unless otherwise noted. Normality was not assumed and nonparametric tests were performed where applicable. Recurrence free and overall survivals were calculated using the Kaplan–Meier method and 95% confidence interval was included for medians and curves. ELISA experiments to access the secretion of GM-CSF from urine were performed in triplicate and the results were reproducible. When enough number of autologous mo-DCs and T cells were obtained from PBMC samples, ELISpot assays were performed in triplicate at the indicated time points.