Combination immunotherapy of solid cancers

Abstract

Methods are provided for treating a solid cancer, particularly solid cancers with an immune-suppressive tumor microenvironment, and notably glioblastoma (GBM). The methods comprise administering a therapeutically effective number of autologous CD34+ hematopoietic stem and progenitor cells (HSPCs) to a patient with a solid cancer, wherein the HSPCs have been engineered to express interferon-a (IFN-α) from an exogenous IFN-α coding sequence that is operably linked to both positive and negative expression control elements that together restrict expression of the exogenous IFN-α coding sequence to differentiated myeloid progeny cells within a tumor microenvironment (TME), and co-administering a therapeutically effective amount of an immune checkpoint inhibitor.

Description

COMBINATION IMMUNOTHERAPY OF SOLID CANCERS1. CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of and priority to U.S. Provisional Application 63 / 632,828, filed on April 11, 2024, the contents of which are incorporated by reference in their entirety.2. BACKGROUND OF THE INVENTION

[0002] Over the past 15 years, a number of immunotherapies, employing a variety of immune-stimulatory approaches, have been approved to treat certain hematological and solid cancers, at times providing unprecedented clinical benefit. Monoclonal antibodies targeting immune checkpoint inhibitors have been approved, for example, to treat triple-negative breast cancer, melanoma, non-small cell lung cancer, head and neck squamous cell cancer, urothelial cancer, microsatellite instability high or a mismatch repair deficient (dMMR) solid tumors, gastric or gastroesophageal junction adenocarcinoma, and renal cell carcinoma, among others. A heterobifunctional soluble T cell receptor (TCR)-targeted T cell engager, tebentafusp, has been approved to treat gp 100-positive melanoma cells in patients with an HLA-A*02:01 haplotype. Antibody-targeted heterobifunctional bispecific T cell engagers have been approved to treat CD 19-positive B-cell precursor acute lymphoblastic leukemia (ALL). Chimeric antigen receptor T cells (CAR-T) have been approved to treat relapsed or refractory multiple myeloma, acute lymphoblastic leukemia (ALL), B-cell precursor ALL, mantle cell lymphoma, and diffuse large B-cell lymphoma (DLBCL).

[0003] Despite notable success in treating hematologic cancers and a select number of solid cancers, immunotherapies have not been demonstrated to be useful in treating all cancers -in particular, certain solid tumors have proven refractory to immune-based treatments, in part due to an immune-suppressive tumor microenvironment (TME) and the absence of a long- lasting response. Glioblastoma multiforme (GBM), the most common and lethal brain tumor, is characterized by a strongly immune-suppressive TME and to date has proven recalcitrant to treatment with immunotherapies. Arrieta et al., J. Clin. Invest. 133 (2): 3163447 (2023).

[0004] Targeted delivery of interferon-a (IFN-a) to the TME to reprogram tumor-associated macrophages (TAMs) toward a proinflammatory phenotype has been proposed as a treatment for solid cancers with strongly immune-suppressive TMEs. Targeted delivery is effected bytransplantation of hematopoietic stem and progenitor cells (HSPCs) that have been transduced ex vivo with a lentiviral vector (LV) expressing IFN-a under positive regulation by the Tie2 enhancer / promoter and negative regulation by HSC-specific microRNAs that recognize complementary target sequences in the expressed IFN-a transcript. Tumorinfiltrating monocyte / macrophage progeny of the transplanted HSPCs release IFN-a in the TME, while IFN-a expression is silenced in undifferentiated hematopoietic stem and progenitor cells and is absent from HSPC progeny that have differentiated into other hematopoietic cell lineages.

[0005] Using a mouse syngeneic GBM model, Naldini and colleagues demonstrated that infusion of such engineered HSPCs leads to release of IFN-a confined to the TME, with robust inhibition of the tumor. Using single cell RNA sequencing (scRNA seq) of GBM- infiltrating hematopoietic cells, the investigators observed reprogramming of tumor- associated macrophages (TAMs) in the murine tumor toward a proinflammatory phenotype, with certain secondary effects on tumor-infiltrating T cells. Birocchi el al., Sci. TransL Med. 14, eabl4106 (2022). While these data are promising, differences between the immune system of inbred mouse strains and humans make it difficult to predict clinical efficacy of immunotherapies from mouse data alone.

[0006] There is a continuing need for immune therapies that are capable of treating GBM and other solid cancers with an immunosuppressive TME.3. SUMMARY OF THE INVENTION

[0007] A non-randomized, open label, phase I / IIa prospective study was conducted involving administration of a single infusion of “Temferon™” - autologous CD34+-enriched HSPCs transduced with a third generation lentiviral vector driving myeloid-specific IFN-a2 expression - to patients affected by glioblastoma (GBM) who have an unmethylated MGMT promoter. Single cell RNA sequencing (scRNAseq) and TCR repertoire analysis were performed on cells from tumor samples obtained from the clinical trial subjects and also from tumor samples of patients outside the trial who had been treated with standard-of-care (SoC) without Temferon™.

[0008] Single cell RNA sequencing (scRNA) analysis showed differential expression between the Temferon™ -treated group and SoC patients without Temferon™, with widespread and strongly significant interferon response and inflammatory response in mostof the cell populations from both the tumor microenvironment (TME) and the GBM tumor cells themselves in the Temferon™ -treated patients.

[0009] In the myeloid cell compartment, a reprogramming toward a pro-inflammatory (Ml- like) phenotype was observed in Temferon™ patients, with (a) a higher proportion of pro- inflammatory macrophages (M ) in the Temferon™ group and (b) a higher proportion of APOC1_2-M in the standard-of-care treated (untreated) group. Comparison of scRNA seq data from post-Temferon™ treatment (second surgery) biopsies of two lesions from the same clinical trial subject, one of which had progressed since first surgery (“Progressed”) and one that had not progressed (“Stable”), showed a noticeable skewing toward an Ml-like phenotype in the Stable lesion as compared to the Progressed lesion, which was conversely enriched in hypoxic (M2 -like) macrophages.

[0010] In addition to the remodeling of the myeloid compartment, a consistent increase in CD8+lymphocytes with effector phenotype was observed in Temferon™ patients as compared to SoC controls, including an increase of effector and effector memory cells as compared to the control dataset.

[0011] ) TCR repertoire analysis of the TCR beta chain (TCRB) performed on the same scRNAseq samples was combined and assessed the presence of the NeoTCR8_ALL gene signature previously identified by Lowery (Lowery et al., Science 375 (6583):877-884 (2022)) and it was observed that the fraction of putatively tumor-reactive cells that are associated with expanded clones are 2-4 fold enriched in Temferon™ -treated patients compared to SoC-treated patients.

[0012] The expression of key molecules such as PD-1, PD-L1 and CTLA4 were analyzed both in the tumor and in TME (myeloid and T cells) in Temferon™-treated and SoC-treated patients, using a discovery dataset of 5 Temferon™ patients and 6 SoC treated patients. When the expression of these markers in T-cells was analyzed, a higher expression of PDCD1 was observed in CD8-Effector and CD8-IFNa populations in Temferon™ patients. Interestingly, most of the PDCD1+T-cells belonged to the stable lesion of the 2-lesions patient. This finding suggests that IFNa promotes the generation of tumor-reactive T cells that are associated with a temporary response, but that the cells upregulate PD1 expression as a consequence of continuous antigen stimulation. Other data suggest that the Temferon™group is more prone to post-tumor reactivity exhaustion, and provides a particular rationale for combining Temferon™ treatment with anti-PDl antibodies.

[0013] These and other findings demonstrate that Ml -biased myeloid reprogramming by Temferon™ leads to activation of CD8 effector T cells and expansion of tumor-reactive T cell clones in the tumor. The upregulation of T cell PD1 expression among these tumor- reactive T cells identifies the PD-1 / PDL1 axis as a target for further increasing the therapeutic efficacy of Temferon™ in GBM, and in other classically immunologically “cold” solid tumors such as renal cell carcinoma, melanoma, high grade osteosarcoma, non-small cell lung cancer (NSCLC), breast cancers, squamous cell carcinoma of the head and neck, bladder cancer, liver and intrahepatic bile duct cancer, gastroesophageal adenocarcinoma / squamous cell cancer, mesothelioma, liver metastases from primary colorectal, breast and urothelial cancers, and melanoma, and epithelial ovarian cancer.

[0014] These findings also have relevance for the treatment of immunological “hot” tumors, such as renal cell carcinoma, melanoma, non-small cell lung cancer (NSCLC), squamous cell carcinoma of the head and neck and bladder cancer. It has previously been shown that a fraction of patients with such immunologically “hot” tumors did not respond to immune checkpoint inhibitors due to a non-functional T cell repertoire. In this regard, Temferon™ - by recruiting and activating newly engaged CD8 effector T cells and tumor-reactive T cells, which will then upregulate PD1 expression - will create a favorable substrate to further enhance the efficacy of immune checkpoint inhibition in such patients.

[0015] Accordingly, methods are presented herein for treating a solid cancer. The methods comprise administering a therapeutically effective number of autologous CD34+hematopoietic stem and progenitor cells (HSPCs) to a patient with a solid cancer, wherein the HSPCs have been engineered to express interferon-a (IFN-a) from an exogenous IFN-a coding sequence that is operably linked to both positive and negative expression control elements that together restrict expression of the exogenous IFN-a coding sequence to differentiated myeloid progeny cells within a tumor microenvironment (TME), and coadministering (prior to, concurrently, and / or subsequent to administration of transduced HSPCs) a therapeutically effective amount of a checkpoint inhibitor.

[0016] One aspect of the present disclosure provides a method of treating a solid cancer, the method comprises administering a therapeutically effective number of autologous CD34+hematopoietic stem and progenitor cells (HSPCs) to a patient with a solid cancer, wherein the HSPCs have been engineered to express interferon-a (IFN-a) from an exogenous IFN-a coding sequence that is operably linked to both positive and negative expression control elements, wherein the control positive and negative control elements together restrict expression of the exogenous IFN-a coding sequence to differentiated myeloid progeny cells within a tumor microenvironment (TME), wherein the therapeutically effective number of autologous CD34+HSPCs is at least 1.5 E6 cells per kilo patient body weight (cells / kg), and administering a therapeutically effective amount of a checkpoint inhibitor.

[0017] In some embodiments, the positive expression control element comprises a Tie2 gene enhancer and promoter operably linked to the exogenous IFN-a coding sequence.

[0018] In some embodiments of any one of the preceding embodiments, the negative expression control element comprises at least one miRNA target sequence (mirT) engineered into the expressed IFN-a transcript, wherein each of the at least one mirT is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells. In some embodiments, the negative expression control comprises a plurality of mirTs engineered into the expressed IFN-a transcript, wherein each of the plurality of mirTs is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells. In some embodiments, at least one mirT is a miR-126 target (miR-126T). In some embodiments, at least one mirT is a miR-130a target (miR-130aT). In some embodiments, the negative control element comprises at least two mir-126T engineered into the IFN-a transcript. In some embodiments, the negative control element consists of two mir-126T in the IFN-a transcript.

[0019] In some embodiments of any one of the preceding embodiments, the HSPCs have been engineered via integration of a lentivirus expression vector. In some embodiments, each of the mirT targets is encoded by a sequence in the integrated lentivirus expression vector that is transcribed into the IFN-a mRNA. In some embodiments, the lentivirus is a replication-defective 3rd-generation pseudotyped vector. In some embodiments, the lentivirus is TIA126-LV.

[0020] In some embodiments of any one of the preceding embodiments, the therapeutically effective number of engineered autologous CD34+HSPCs is at least 2.0 E6 cells / kg.

[0021] In some embodiments of any one of the preceding embodiments, the therapeutically effective number of engineered autologous CD34+HSPCs is at least 2.5 E6 cells / kg.

[0022] In some embodiments of any one of the preceding embodiments, the therapeutically effective number of engineered autologous CD34+HSPCs is at least 3.0 E6 cells / kg.

[0023] In some embodiments of any one of the preceding embodiments, the therapeutically effective number of engineered autologous CD34+HSPCs is at least 3.5 E6 cells / kg.

[0024] In some embodiments of any one of the preceding embodiments, the therapeutically effective number of engineered autologous CD34+HSPCs is at least 4.0 E6 cells / kg. In some embodiments, the therapeutically effective number of engineered autologous CD34+HSPCs is 4E6 cells / kg.

[0025] In some embodiments of any one of the preceding embodiments, the method further comprises administering untransduced CD34+supporter cells in combination with the engineered CD34+HSPCs. In some embodiments, 3E6 supporter cells / kg patient body weight are administered. In some embodiments, 2E6 supporter cells / kg are administered.

[0026] In some embodiments of any one of the preceding embodiments, the method further comprises the step, preceding administration of engineered autologous CD34+cells, of administering at least one sub-myeloablative conditioning regimen. In some embodiments, the at least one sub-myeloablative conditioning regimen is selected from BCNU and thiotepa, busulfan and thiotepa, and busulfan without thiotepa. In some embodiments, the at least one sub-myeloablative conditioning regimen is busulfan without thiotepa.

[0027] In some embodiments of any one of the preceding embodiments, the solid cancer is glioblastoma multiforme (GBM). In some embodiments, the solid cancer is GBM with unmethylated MGMT promoter (uMGMT). In some embodiments, the method further comprises the step, prior to administering the engineered autologous CD34+HSPCs, of surgically resecting the tumor mass. In some embodiments, the method further comprises the step, before and / or after surgical resection and before administering the engineered autologous CD34+HSPCs, of administering radiotherapy to the patient.

[0028] In some embodiments of any one of the preceding embodiments, the checkpoint inhibitor is an inhibitor of PD-1, PD-L1, CTLA-4, LAG-3, ICOS, BTLA, TIM-3, TIGIT, or NKG2A. In some embodiments, the checkpoint inhibitor is an antibody or antigen-binding fragment that binds specifically to PD-1, PD-L1, CTLA-4, LAG-3, ICOS, BTLA, TIM-3, TIGIT, or NKG2A. In some embodiments, the checkpoint inhibitor specifically binds to PD-1. In some embodiments, the PD-1 inhibitor is an anti-PD-1 antibody selected from the group consisting of: pembrolizumab, nivolumab, and cemiplimab. In some embodiments, the anti- PD-1 antibody is pembrolizumab. In some embodiments, the anti-PD-1 antibody is nivolumab. In some embodiments, the anti-PD-1 antibody is cemiplimab. In some embodiments, the checkpoint inhibitor specifically binds to PD-L1. In some embodiments, the PD-L1 inhibitor is an anti-PD-Ll antibody selected from the group consisting of: atezolizumab, avelumab, and durvalumab. In some embodiments, the checkpoint inhibitor specifically binds to CTLA-4. In some embodiments, the CTLA-4 inhibitor is an anti-CTLA- 4 antibody selected from the group consisting of: ipilimumab and tremelimumab. In some embodiments, the checkpoint inhibitor specifically binds to LAG-3. In some embodiments, the LAG-3 inhibitor is the anti-LAG-3 antibody is relatlimab. In some embodiments, the checkpoint inhibitor specifically binds to ICOS. In some embodiments, the anti-ICOS antibody is MEDI-570. In some embodiments, the checkpoint inhibitor specifically binds to BTLA. In some embodiments, the checkpoint inhibitor specifically binds to TIM-3. In some embodiments, the checkpoint inhibitor specifically binds to TIGIT. In some embodiments, the anti-TIGIT antibody is selected from vibostolimab, etigilimab, and tiragolumab. In some embodiments, the checkpoint inhibitor specifically binds to NKG2A.

[0029] In some embodiments of any one of the preceding embodiments, the checkpoint inhibitor is administered prior to, concurrently with, and / or subsequent to administration of the autologous CD34+HSPCs. In some embodiments, the checkpoint inhibitor is first administered after confirmation of HSPC engraftment.4. BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0030] These and other features, aspects, and advantages of the present invention will become better understood with regard to the following description, and accompanying drawings, where:

[0031] FIG. 1 is a circular map of the p.INFa plasmid used in the manufacture of Temferon™;

[0032] FIG 2 illustrates salient treatment characteristics as well as current disposition of clinical trial subjects for Cohorts 1-8 in the clinical trial described in Example 1 (NCT03866109);

[0033] FIG. 3 presents Gene Set Enrichment analysis performed on ranked log-fold change gene lists obtained from intra-population differential expression analysis between the Temferon™-treated group and subjects outside the clinical trial receiving Standard-of-Care (SoC) without Temferon™, showing a widespread and strongly significant (NES > 0; p.adj(BH) < 0.01) interferon response and inflammatory response in most of the cell populations from both the tumor microenvironment (TME) and the GBM tumor cells themselves in the Temferon™ -treated patients.

[0034] FIG. 4A illustrates myeloid TME reprogramming to Ml -like populations from M2- like populations in the Temferon™-treated patients, with (a) a higher proportion of pro- inflammatory macrophages (M) in the Temferon™ group and (b) a higher proportion of APOC1_2-M in the standard-of-care treated (untreated) group. FIG. 4B presents superimposed UMAP plots of scRNA seq data from post-Temferon™ treatment (second surgery) biopsies of two lesions from the same Temferon™-treated subject, one of which had progressed since first surgery (“Progressed”) and one that had not progressed (“Stable”), showing a noticeable skewing toward an Ml -like phenotype in the Stable lesion as compared to the Progressed lesion, which was conversely enriched in hypoxic (M2-like) macrophages.

[0035] FIG. 5A is a heat map showing levels of expression of various T cell markers (y-axis) in tumor samples drawn from individual subjects in the Temferon™ clinical trial and in tumor samples from non-trial subjects with primary, recurrent, or unknown stage tumor status who had not been treated with Temferon™. Temferon™-treated patients are indicated with arrows. Tumors from the Temferon™-treated group showed an increased proportion of CD8 cells. FIG. 5B quantifies CD8 subset data from Temferon™-treated patients (“TEMFERON”) and the Gbmap-core control dataset subjects who did not receive Temferon™ (“UNTREATED”), with increasing levels of statistical significance indicated as n.s. (not significant), *, **, ***, and ****. These data show a statistically significant increase in Temferon™ -treated patients in CD8 lymphocytes with an effector phenotype and an activated phenotype. Only patients with > 50 cells were included.

[0036] FIG. 6A is a UMAP plot showing increased presence of T cells that are putatively tumor-reactive based on a transcriptional signature previously identified by Lowery et cd.. Science 375 (6583):877-884 (2022) (“Lowery”) in tumor biopsies of Temferon™ -treated patients as compared to Temferon™-naive standard-of-care controls. FIG. 6B overlays topographic map lines showing expanded clonotypes. The fraction of putatively tumor-reactive cells that are associated with expanded clones are 2-4 fold enriched in Temferon™- treated patients compared to SoC-treated patients.

[0037] FIGs. 7A-7C show distribution of gene expression levels of PD-L1 (CD274) (FIG. 7 A), CTLA4 (FIG. 7B) and PD-1 (PDCD1) (FIG. 7C) across CD4_Treg and CD8-Effector- like populations among 5 Temferon™-treated and 6 SoC-treated (“UNTREATED”) patients. Values are expressed as loglp.

[0038] FIG. 8 is a boxplot of frequency ratios between PD I CTLA4 and PD1 CTLA4+CD8-Effector T-cells. The median value of the Temferon™ group is ~2-fold the one in SoC treated group.

[0039] FIG. 9 presents UMAP plots of scRNA seq data from CD8+ T cells that belong to clonotypes that are “private” to second surgery samples - that is, not also present in peripheral blood (PB-associated samples) - in the progressed lesion (left panel) and in the stable lesion (right panel) of the patient who had two lesions sampled in a second surgery. In light-blue are cells that are defined as expanded (that is, in a clonotype with frequency > 0.1%) tumor-reactive TILs (trTILs) according to the Lowery et al. NeoTCR8_ALL signature. The quota of trTILs is much higher in the stable versus the progressed lesion.

[0040] FIG. 10A shows a schematic of the design of the phase I / IIa clinical study in patients with newly diagnosed glioblastoma multiforme and non-mutated isocitrate dehydrogenase gene. MGMT: 06-methylguanine-DNA methyltransferase. *: Day-6 for patients receiving double alkylator conditioning, Day-4 for patients receiving Busulfan monotherapy.

[0041] FIG. 10B is a table describing the 8 cohorts (3 patients each) according to conditioning regimen, Temferon™ target doses, Supporter CD34+cell target doses, as well as the fraction of transduced CD34+cells actually infused in the patients (input fraction) calculated as the (number of CD34+cells in the thawed and reconstituted Temferon™ product) x (transduction efficiency measured on clonogenic progenitors) / (nominal amount of CD34+cells in the Supporters + nominal numbers of cells in the reconstituted Temferon™ product).

[0042] FIG. 10C is a line plot that shows the engraftment of transduced myeloid progenitors in the bone marrow (BM) at +30, +90, +180 and +360 days from Temferon™ infusion (dark gray circles, squares, and triangles) estimated from the vector copy number (VCN) transplantas a percentage of total BM CD33+cells, as compared to the input fraction of transduced CD34+cells (black circles).

[0043] FIG. 10D depicts line plots separately for each cohort showing vector copy number (VCN) in CD14+peripheral blood mononuclear cells (PBMC) as a function of time post infusion (patient codes indicated as TEM-GXX).

[0044] FIG. 10E depicts line plots separately for each cohort showing interferon-alpha (IFN- a) concentrations in the plasma measured by ELISA.

[0045] FIG. 10F shows (solid lines) overall survival (OS, top graph) and progression-free survival (PFS, bottom graph) from diagnosis (date of initial surgery). Dotted lines in the graphs indicate the 95% confidence interval. In the OS graph, the median survival is around 17 months; in the PFS, the median survival is around 9 months.

[0046] FIG. 10G shows a swim lane plot according to overall survival.

[0047] FIG. 11A depicts, separately for each cohort, line plots showing VCN in peripheral blood cell subpopulations for individual patients (TEM-GXX). PBMC, peripheral blood mononuclear cells. CD15+cells, granulocytes. VCN was estimated (est.) for lymphoid cells, as follows: est. VCN (lymphoid) = (VCNPBMC - VCNCD14+ x FCD14+) / (1-FCD14+), whereby FCD14+ is the estimated fraction of CD14+monocytes within PBMCs (absolute monocyte count / (ab solute monocyte count + absolute lymphocyte count) at the given timepoint.

[0048] FIG. 11B depicts line plots for each cohort showing VCN in bone marrow mononuclear cells (BM-MNC) or BM CD33+myeloid cells.

[0049] FIG. 11C depicts lines plots showing engraftment of TIA126 LV-transduced clonogenic progenitors in the BM at the indicated timepoints post Temferon™ infusion. The symbols indicate the number of progenitor-derived colonies (CFU-G / M or BFU / CFU-E) tested positive for the LV (VCN>0.5) divided by the total number of analyzed colonies. Error bars denote the 80% confidence interval calculated according to the method of Clopper and Pearson. An average of 35 individual CFU-G / M (minimum: 16) and 47 individual BFU / CFU-E (minimum: 38) were screened per patient and timepoint.

[0050] FIG. 11D are bar graphs showing VCN distribution within TIA126 LV+ (VCN>0.5) colony-forming units (CFUs) from the drug products from all patients (left graph, total CFUs and split for erythroid BFU-E and myeloid CFU-G / M) and from BM CD34+progenitors isolated at the indicated timepoints post Temferon™ infusion (right graph, shown for the total CFUs). The number above the bars indicate the number of individual TIA126 LV+ colonies analyzed. VCN 1 (0.5-1.4); VCN 2 (1.5-2.4); VCN 3 (2.5-3.4); VCN 4 (3.5-4.4); VCN5+ (4.5-20); VCN values >20 were rejected.

[0051] FIG. HE are graphs showing VCN in CD14+monocytes (right graph) and shortlived, CD15+granulocytes (left graph) at 2, 3, 4, and 6 months post Temferon™ infusion, normalized to the first in vivo datapoint at 1 month for each patient. C / T: Carmustin (BCNU) + Thiotepa conditioning; Bu / T: Busulfan + Thiotepa; Bu: Busulfan monotherapy.

[0052] FIG. 12A shows a uniform manifold approximation and projection (UMAP) plot with internal dataset embedding (7 post-Temferon™ and 6 control samples) with annotated T cell populations from extended dataset (n = 37,392 cells).

[0053] FIG. 12B shows a UMAP plot with clonotype class according to the TCR clonotype frequency (based on TCRB amino acid sequence) for each single cell within the internal TIL dataset. H, hyperexpanded (10-100%); L, large (1% to 10%); M, medium (0.1% - 1%); S, small (0.01- 0.1%); no, cells where no TCR sequences were amplified (n = 12,549).

[0054] FIG. 12C shows a UMAP plot mapping the predicted tumor-reactive TILs (trTILs) based on NeoTCR8 signature (AddModule Score - Ucell) onto the UMAP embedding encompassing only CD8 T cells with a TCRB sequence.

[0055] FIG. 12D shows an intra-cluster gene set enrichment analysis (GSEA). Heatmap with normalized enrichment scores values (NES) for statistically significant MSigDB Hallmark terms present in at least 2 populations. GSEA was performed according to a pre-ranked logFC gene list based on DEGs between Temferon™ vs. Controls in each population. Light gray tiles: not significant (p.adj > 0.01; BH correction); dark gray tiles: significantly positive or negative enriched terms.

[0056] FIG. 12E shows an interaction analysis between Myeloid populations and GZMK+CD8 Effector / Exhausted (Eff / Exh) T cells population. Heatmaps show top interacting receptors with 30 prioritized ligands (left), the log fold change (LFC) of the respective ligandin Temferon™ vs control samples (center) and interaction potential scores between top 30 prioritized ligands vs downstream targets (right).

[0057] FIG. 13 shows the adverse events in patients during and within 24 hours of Temferon™ infusion.

[0058] FIG. 14A are line plots showing hematologic recovery by cohort. Absolute blood counts of neutrophils (ANC, left graph) and platelets (PLT, right graph) during the first 90 days from Temferon™ infusion are shown (median, n=3 per cohort), as determined by routine automated complete blood count in the hospital. Arrows indicate slower kinetics of ANC and PLT decrease in the cohorts receiving busulfan-only conditioning. ULN, upper limit of normal; LLN, lower limit of normal.

[0059] FIG. 14B summarizes neutrophil and platelet engraftment, defined as the first of 3 consecutive days of ANC >0.5xl09 / L and PLT >20xl09 / L, in the absence of transfusion for at least 7 consecutive days.

[0060] FIG. 14C tabulates the description and incidence of severe adverse events (SAE) according to subjects, timing (before or after day +90 from Temferon™ infusion), cohort and conditioning regimen. * SAE rate by conditioning was calculated as the number of events, divided by the number of patients assigned to one of the 3 conditioning regimens: C / T: Carmustin+Thiotepa, n=15; Bu / T: Busulfan+Thiotepa, n=3; Bu: Busulfan monotherapy, n=6.

[0061] FIG. 14D shows the interferon-alpha (IFN-a) levels in the cerebrospinal fluid (CSF) in the patients after Temferon™ administration, as determined by high-sensitivity ELISA, scr: baseline levels at screening. Negative values indicate undetectable levels of IFN-a.

[0062] FIG. 14E shows timepoints of tumor sampling after Temferon™ infusion. Black boxes denote time of radiologically-determined progressive disease (PD), while arrows and numbers denote the time of second surgery, counted in days from Temferon™ infusion.

[0063] FIG. 14F summarizes biomarkers of Temferon™ activity in the tumor. Vector copy number (VCN) was assessed either on formalin-fixed paraffin-embedded (FFPE) tissue samples, or on the CD45+hematopoietic tumor infiltrate. <LLOQ, value below the lower limit of quantification, n / a, data point not available due to insufficient material. The % of TIE2+ tumor associated macrophages (TAMs) was measured by flow cytometry. Interferonresponse genes (OAS1, IFIT1, IRF7) were measured by quantitative PCR on FFPE, indicated as the 2'AACtvalues between the second surgery and first surgery sample from a given patient (fold change to diagnosis, fc to Dx).

[0064] FIG. 15A are line plots showing hematologic recovery by cohort. Absolute blood count of lymphocytes (ALC), absolute blood count of monocytes (AMC), and hemoglobin (HGB) levels during the first 90 days from Temferon™ infusion are shown (median, n=3 per cohort), as determined by routine automated complete blood count in the hospital. Arrow indicates stability of ALC in the cohorts receiving busulfan-only conditioning. ULN, upper limit of normal; LLN, lower limit of normal.

[0065] FIG. 15B are line plots showing absolute neutrophil counts (ANC), absolute monocyte counts (AMC), absolute lymphocyte counts (ALC), as well as hemoglobin levels (HGB) and platelet counts (PLT) in the blood during long-term follow up after Temferon™ administration. ULN, upper limit of normal; LLN, lower limit of normal.

[0066] FIG. 15C are line plots showing lymphocyte subtyping by multicolor flow cytometry in the blood after Temferon™ administration. Shown are the absolute counts of CD4+T helper cells, CD8+cytotoxic T lymphocytes, CD19+B lymphocytes, natural killer (NK) cells (CD3 CD16+or CD56+within the lymphocyte gate), and CD4+CD25hlCD127‘ regulatory T cells (Tregs).

[0067] FIG. 16 shows all of the treatment-emergent adverse events during the phase Ulla clinical trial.

[0068] FIG. 17A shows the distribution of lentiviral integration sites across the human genome for each single patient following Temferon™ administration. Lentiviral integration sites were determined by Sonication Linker-mediated-PCR (SLiM-PCR) from longitudinal blood and bone marrow samples from patients that had received Temferon™. Note that the density peaks of integration sites corresponds well with gene density across the various chromosomes for all patients.

[0069] FIG. 17B shows an analysis of common integration sites. No statistically significant overrepresentation of integrations into known oncogenes or tumor suppressor genes was detected in any patient.

[0070] FIG. 17C are line plots showing clonal diversity within CD15+blood granulocytes (upper graph) or CD33+bone marrow myeloid precursors (lower graph) over time, expressed as Shannon index. Note that clonal diversity is decreasing over time, in line with the decreasing VCN.

[0071] FIG. 18 shows the plot cytometry gating for cells isolated from dissociated GBM tumors from trial patients. FIG. 18A shows flow cytometric analysis of CD45+-enriched cells from dissociated GBM tumors. The gating strategy to identify TIE2-expressing monocytes / macrophages (TEMs) is shown for a representative sample. Note that TIE2 staining is limited to CD14+HLA-DR+myeloid cells, gates for TIE2+ staining were set based on an isotype control (background <1% within the CD14+or CD14+DR+gate). FIG. 18B shows the expression of CD83, a molecule involved in antigen presentation and associated with inflammatory macrophages, on total CD14+cells or TEMs. FIG. 18C shows representative histograms showing TIE2 (TEK) expression (black) as compared to isotype control (grey) on CD14+HLA-DRhlghtumor-infiltrating myeloid cells. FIG. 18D shows quantification of TEMs within the total CD14+HLA-DRhlghtumor-infiltrating myeloid cell population (n=8 samples, 4 from patients post Temferon™ treatment, and 4 from control GBM patients). Box and whisker plot shows the median, interquartile range and min / max values, each datapoint represents a sample.

[0072] FIG. 19A are graphs showing transcript expression of 3 representative interferon response genes (IRGs) relative to the Hypoxanthine Phosphoribosyltransferase 1 (HPRT) housekeeping gene measured by digital droplet PCR in peripheral blood (PB) or bone marrow mononuclear cell (BM-MNC) subpopulations at screening, e.g., before Temferon™ treatment (baseline expression, left graphs, for each IRG), as well as in bulk tumor (right graphs, for each IRG) at first surgery (GBM-pre Temferon™) or at second surgery (GBM- post Temferon™). Each dot represents a patient, the line indicates the median. Note the population-intrinsic differences in baseline IRG expression. OAS1, Oligoadenylate Synthetase 1 (top graph); IFIT1, Interferon Induced Protein With Tetratricopeptide Repeats 1 (middle graph); IRF7, Interferon Regulatory Factor 7 (bottom graph).

[0073] FIG. 19B are line graphs showing relative change of IRG expression in peripheral blood or bone marrow subpopulations at the indicated timepoint (days) after Temferon™ infusion, grouped by subpopulation and cohort. Shown is the median fold change of the OAS1 / HPRT, IFIT1 / HPRT and IRF7 / HPRT ratios with respect to baseline (value atscreening, see A) for individual patients. Error bars indicate the range for the relative change of the 3 IRGs.

[0074] FIG. 20A is a UMAP projection of the merged object from the CD45+-enriched tumor dissociates from these 12 samples (internal dataset). Single cell RNA sequence of GBM was performed on n=6 patients after Temferon™ infusion, as well as n=6 recurrently GBM tumors from patients treated with stand or care (SoC). The insert shows the number and proportion of major cell types and lineages contained in the dataset.

[0075] FIG. 20B shows unsupervised clustering (resolution: 1.8) of the tumor-associated myeloid compartment (left) and provisional classification (right) based on marker gene expression.

[0076] FIG. 20C shows projection of the refined myeloid population labels from the GBM single cell atlas on the internal dataset.

[0077] FIG. 20D shows unsupervised clustering at ultra-high resolution (res 4) of the tumor- associated T- / NK cell compartment (left) and provisional classification (right) based on marker gene expression.

[0078] FIG. 20E shows projection of the refined TIL population labels from the GBM single cell atlas on the internal dataset.

[0079] FIG. 21A shows a UMAP plot with annotation of the myeloid cell compartment inside the GBM microenvironment (n=69 samples, ~ 150k cells) of single cell GBM tumor microenvironment landscapes (GBM atlas) assembled from scRNAseq data from 7 post- Temferon™ samples, 6 control samples at recurrence (internal dataset), and multiple published datasets from the Core GBmap atlas.

[0080] FIG. 21B is a dot plot with the most representative marker genes in the annotated myeloid TME cell populations. The size of dot shows the percentage of cells expressing the gene (column) in the cell population (row); shade represents the z-score expression of the gene across populations.

[0081] FIG. 21C is a heatmap of population frequencies within each patient with unsupervised clustering of the samples into 4 myeloid macro-clusters (M-I to M-IV).

[0082] FIG. 21D is a box plot with frequency distribution of inflammatory macrophages, MIF+ macrophages, and APOC+ macrophages in Temferon™ -treated patients vs. controls. Each dot is a single patient (n=69).

[0083] FIG. 21E is a UMAP plot with annotation of the tumor infiltrating lymphocyte (TIL) compartment (n=38 samples, ~ 80k cells).

[0084] FIG. 21F is a dot plot with the most representative marker genes identified in the annotated T cells populations. The size of dot shows the percentage of cells expressing the gene (column) in the cell population (row); shade represents the z-score expression of the gene across populations.

[0085] FIG. 21G is a heatmap of population frequencies within each patient with unsupervised clustering of the samples highlighting 2 distinct T cell macro-clusters (T-I, T- II).

[0086] FIG. 21H presents boxplots with frequency distribution of T cell populations that were significantly different between Temferon™ and controls. Frequency-dependent charts (C, D, G and H) and statistics were produced based on patients contributing at least 100 cells. The heatmaps show column-scaled values of the number of cells in each population within each donor. (*) denotes a statistically significant difference in population frequency (p < 0.05) between Temferon™ and controls: Wilcoxon rank sum test (* p<0.05; ** p<0.01; empty; p > 0.05). SoC: patients treated according to standard of care.

[0087] FIG. 22A shows a UMAP plot with annotation of the non-hematopoietic (CD45n) cell compartment composed mainly of tumor cells (61 patients n = 142,173 cells).

[0088] FIG. 22B is a dot plot with the 3 most representative marker genes in the annotated CD45negtumor cell populations.

[0089] FIG. 22C is a heatmap of cell type populations frequencies within each patient (n = 61) with unsupervised clustering into 4 GBM (G) macro-clusters (G-I to G-IV). Frequency dependent chart and associated statistics were produced based on patients contributing at least 100 cells. The heatmap shows column-scaled values of the number of cells in each population within each donor. SoC: patients treated according to standard of care.

[0090] FIG. 22D is a box and whisker plot showing the relative frequency (% within tumor- associated myeloid cells) of each annotated cluster within a GBM sample post Temferon™ treatment or from patients who have received other treatments. Each dot represents an individual sample, whereas box and whisker plots show the group’s median with interquartile range and minimum / maximum values. Statistical comparison between patients that have received Temferon™ and patients that have received other treatments was performed by Wilcoxon log rank test (* p<0.05; ** p<0.01; ns, not significant).

[0091] FIG. 22E is a box and whisker plot showing the relative frequency (% within TILs) of each annotated cluster within a GBM sample post-Temferon™ treatment or from patients who have received other treatments. Each dot represents an individual sample; box and whisker plots show the group’s median with interquartile range and minimum / maximum values. Statistical comparison between patients that have received Temferon™ and patients that have received other treatments was performed by Wilcoxon log rank test (* p<0.05; ** p<0.01; ns, not significant).

[0092] FIG. 23 is a plot showing dynamics of the TIL clonotypes identified in the Temferon™ patients by scTCRseq. Around half (46.9%) of the TIL clonotypes identified by scTCRseq in the recurrent tumor at second surgery could be traced back in a bulk TCR sequencing dataset comprising the primary tumor at diagnosis, peripheral blood T cells before and during several time points after Temferon™ administration, and an independent sampling from the relapsed tumors. Each row refers to a unique TCRB clonotype identified in both the scTCRseq and bulk TCRseq datasets. The first column on the left shows patient distribution, which is well balanced across the TCR clonotypes. The second column shows the clonotype size in the scTCRseq dataset from the recurrent tumor. The block with the 6 columns on the right shows TCR clonotype frequency in the indicated sampling source and timepoint. About half of the clonotypes were present with high clonal frequency in the blood, and these generally had low clonal size in the tumor, consistently between the scTCRseq and clonoSEQ data, suggesting that they were bystanders rather than tumor-reactive clones. On the other hand, around 40% of the clonotypes were present at high frequency in the relapsed tumor but low frequency in the blood. Interestingly, some of them were already enriched in the tumor at diagnosis and persisted at low levels in the blood throughout all timepoints, while others emerged over time, after day 90, or were private to the relapsed tumor.

[0093] FIG. 24A shows radiologic tumor progression in patient TEM-G11. The timeline indicates the different lines of treatment that the patient had received in between the MRI assessments. WBRT: whole brain radiotherapy. The patient had 4 target lesions (TL), the volume of which is shown over time. The 3 largest TLs are shown in the sagittal crosssections. White dotted circles, contrast-enhancing tumor. Radiologic images also show tumors and perilesional edema.

[0094] FIG. 24B is a box graph showing expression of interferon response genes (OAS1, IFIT1, IRF7; fold change over HPRT) measured by digital droplet PCR in biopsies from G11-1 (TL1, stable lesion, black) or G11-2 (TL2, progressing lesion, dark gray).

[0095] FIG. 24C are plots showing TIE2 surface expression (flow cytometric) in tumor- associated macrophages (CD45+CD14+HLA-DR+) from TEM-G11-1 (TL1, stable lesion) or TEM-G11-2 (TL2, progressing lesion), as compared to isotype staining.

[0096] FIG. 24D are UMAP plots with simplified annotation of the macrophage compartment from TEM-G11-1 (TL1, stable lesion) and TEM-G11-2 (TL2, progressing lesion), as defined by the extended scRNAseq myeloid dataset, and relative population distribution in the 2 lesions.

[0097] FIG. 24E are plots showing blood perfusion of TL1 and TL2 measured as standardized cerebral blood volume (sCBV) in enhancing tumor tissue by functional MRI.

[0098] FIG. 24F are bar graphs showing the proportion of CD8+T cells with a NeoTCR8_ALL signature above (gray, trTILs) or below (light gray, bystander T cells) a threshold defined as the minimum value of the top scoring cluster.

[0099] FIG. 24G is a volcano plot with differentially expressed genes between trTILs and by-stander T cells from the stable TL1, using T cell subpopulation as a covariate. The top 15 differentially expressed genes according to logFC are shown; black dots represent statistically significant up and down regulated DEGs.

[0100] FIG. 24H is a heat plot showing putative trTILs defined as T cells with a high neoTCR8 score in scRNAseq, a TCRB sequence matched to each individual neoTCR8high cell in scTCRseq and absence / low abundance (< 0.01%) of this TCRB sequence in scTCRseq performed on peripheral blood T cells at the time of surgery. The 4 columns on the left showthe size (class) and the relative representation (frequency) of each individual trTIL clonotype within TL1 (stable lesion) or TL2 (progressing lesion), as assessed by scTCRseq. The heatmap (right) shows clonotype dynamics by backtracking these trTIL clonotypes in longitudinal bulk TCR sequencing data across therapy timepoints in blood and tumor (1st and 2nd surgery). Values represent the trTILs clonotype (rows) frequencies obtained by clonoSEQ, scaled to the maximum value along timepoints (columns). Unsupervised clustering of the bulk TCR sequencing data identified 4 trTIL macro-clusters (I-IV) with different clonal dynamics in blood and tumor.

[0101] FIG. 241 is a volcano plot with differentially expressed genes between trTILs from the stable TL1 and the progressing TL2, using TCRB clonotype as a covariate. The top 15 differentially expressed genes according to logFC are shown with labels; black dots represent statistically significant up and down regulated DEGs.

[0102] FIG. 25A is a UMAP plot mapping the NeoTCR8 signature onto the T cell-focused UMAP plot with internal dataset embeddings for patients after treatment with Temferon™ or not (SoC).

[0103] FIG. 25B is an alluvial plot of TCRB clonotypes associated with a high NeoTCR8 signature (trTILs, gray) or not (by-standers, light gray), with respect to sample origin (patient after Temferon™ treatment); MLGX: control GBM sample at recurrence upon standard of care treatment), T cell subcluster and clone size.

[0104] FIG. 26 shows a UMAP plot mapping the Tie2+ signature in the TEM in trial patient TEM-G11.

[0105] FIG. 27 shows a Kaplan-Meier overall survival (OS) analysis for percentage of CD8 expression in trial patients at second surgery. The numbers in the bottom part of the figure are “number at risk.”5. DETAILED DESCRIPTION OF THE INVENTION5.1. Combination immunotherapy of solid cancers

[0106] In a first aspect, methods are provided for treating a solid cancer. The method comprises administering a therapeutically effective number of autologous CD34+hematopoietic stem and progenitor cells (HSPCs) to a patient with a solid cancer, wherein the HSPCs have been engineered to express interferon-a (IFN-a) from an exogenous IFN-a coding sequence that is operably linked to both positive and negative expression control elements that together restrict expression of the exogenous IFN-a coding sequence to differentiated myeloid progeny cells within a tumor microenvironment (TME); and coadministering (either prior to, concurrently with, and / or subsequent to administration of the autologous CD34+HSPCs) a therapeutically effective amount of a checkpoint inhibitor.

[0107] In various embodiments, the positive expression control element comprises a Tie2 gene enhancer and promoter operably linked to the exogenous IFN-a coding sequence.

[0108] In some embodiments, the negative expression control element comprises at least one miRNA target sequence (mirT) engineered into the expressed IFN-a transcript, wherein each of the at least one mirT is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells. In certain embodiments, the negative expression control comprises a plurality of mirTs engineered into the expressed IFN-a transcript, wherein each of the plurality of mirTs is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells. In certain embodiments, at least one mirT is a miR-126 target (miR- 126T). In certain embodiments, at least one mirT is a miR-130a target (miR-130aT). In currently preferred embodiments, the negative control element comprises at least two mir- 126T engineered into the IFN-a transcript. And in a specific embodiment, the negative control element consists of two mir-126T in the IFN-a transcript.

[0109] In various embodiments, the HSPCs have been engineered via integration of a lentivirus expression vector. In some embodiments, each of the mirT targets is encoded by a sequence in the integrated lentivirus expression vector that is transcribed into the IFN-a mRNA. In some embodiments, the lentivirus is a replication-defective 3rd-generation VSV-G pseudotyped vector. In specific embodiments, the lentivirus is TIA126-LV.

[0110] In various embodiments, the therapeutically effective number of autologous CD34 HSPCs is at least 1.5 x 106(1.5 E6) cells per kilo patient body weight (cells / kg). In someembodiments, the therapeutically effective number of autologous CD34+HSPCs is 2 x 106(2E6) cells per kilo patient body weight (cells / kg). In some embodiments, the therapeutically effective number of autologous CD34+HSPCs is at least 3E6 cells / kg. In some embodiments, the therapeutically effective number of autologous CD34+HSPCs is greater than 3E6 cells / kg. In some embodiments, the therapeutically effective number of autologous CD34+HSPCs is 4E6 cells / kg. In some embodiments, the therapeutically effect number of autologous CD34+HSPCs is 4.4 E6 cells / kg.[OHl] In some embodiments, the methods further comprise administering untransduced CD34+supporter cells in combination with the engineered CD34+HSPCs. In certain embodiments, 3E6 supporter cells / kg patient body weight are administered. In certain embodiments, 2E6 supporter cells / kg are administered.

[0112] In some embodiments, the method further comprises the step, preceding administration of engineered autologous CD34+cells, of administering at least one sub- myeloablative conditioning regimen.

[0113] In certain embodiments, the at least one sub-myeloablative conditioning regimen is selected from BCNU and thiotepa, busulfan and thiotepa, and busulfan without thiotepa. In specific embodiments, the at least one sub-myeloablative conditioning regimen is busulfan without thiotepa. In particular embodiments, a single sub-myeloablative conditioning regimen is administered. In specific embodiments, the single sub-myeloablative conditioning regimen is busulfan without thiotepa.

[0114] In various embodiments, the solid tumor is selected from high grade glioma, glioblastoma, renal cell carcinoma, melanoma, high grade osteosarcoma, non-small cell lung cancer (NSCLC), breast cancers, squamous cell carcinoma of the head and neck, bladder cancer, liver and intrahepatic bile duct cancer, gastroesophageal adenocarcinoma / squamous cell cancer, mesothelioma, liver metastases from primary colorectal, breast and urothelial cancers, melanoma, and epithelial ovarian cancer. In various embodiments, the solid tumor is selected from renal cell carcinoma, melanoma, non-small cell lung cancer (NSCLC), squamous cell carcinoma of the head and neck, and bladder cancer.

[0115] In some embodiments, the solid cancer is glioblastoma multiforme (GBM). In certain embodiments, the solid cancer is GBM with unmethylated MGMT promoter (uMGMT).

[0116] In particular embodiments, the method of treating GBM further comprises the step, prior to administering the CD34+engineered HSPCs, of surgically resecting the tumor mass. In some embodiments, the method of treating GBM further comprises the step, before and / or after surgical resection and before administering the CD34+engineered HSPCs, of administering radiotherapy to the patient.

[0117] In various embodiments of the methods presented herein, the checkpoint inhibitor is an inhibitor of PD-1, PD-L1, CTLA-4, LAG-3, ICOS, BTLA, TIM-3, TIGIT, or NKG2A. In typical embodiments, the checkpoint inhibitor is an antibody or antigen-binding fragment thereof that binds specifically to PD-1, PD-L1, CTLA-4, LAG-3, ICOS, BTLA, TIM-3, TIGIT, or NKG2A.

[0118] In certain embodiments, the checkpoint inhibitor specifically binds to PD-1. In specific embodiments, the PD-1 inhibitor is an anti -PD-1 antibody selected from the group consisting of: pembrolizumab, nivolumab, and cemiplimab.

[0119] In certain embodiments, the checkpoint inhibitor specifically binds to PD-L1. In specific embodiments, the PD-L1 inhibitor is an anti-PD-Ll antibody selected from the group consisting of: atezolizumab, avelumab, and durvalumab.

[0120] In certain embodiments, the checkpoint inhibitor specifically binds to CTLA-4. In specific embodiments, the CTLA-4 inhibitor is an anti-CTLA-4 antibody selected from the group consisting of: ipilimumab and tremelimumab.

[0121] In certain embodiments, the checkpoint inhibitor specifically binds to LAG-3. In specific embodiments, the anti-LAG-3 antibody is relatlimab.

[0122] In certain embodiments, the checkpoint inhibitor specifically binds to ICOS. In specific embodiments, the anti-ICOS antibody is MEDI-570.

[0123] In certain embodiments, the checkpoint inhibitor specifically binds to BTLA.

[0124] In certain embodiments, the checkpoint inhibitor specifically binds to TIM-3.

[0125] In certain embodiments, the checkpoint inhibitor specifically binds to TIGIT. In specific embodiments, the anti-TIGIT antibody is selected from vibostolimab, etigilimab, and tiragolumab.

[0126] In certain embodiments, the checkpoint inhibitor specifically binds to NKG2A.

[0127] In some embodiments, the checkpoint inhibitor is administered prior to, concurrently with, and / or subsequent to administration of the autologous CD34+HSPCs. In certain embodiments, the checkpoint inhibitor is administered after confirmation of HSPC engraftment.

[0128] One aspect of the present disclosure provides a method of treating glioblastoma multiforme (GBM), the method comprising administering a therapeutically effective number of autologous CD34+hematopoietic stem and progenitor cells (HSPCs) to a patient with GBM, wherein the HSPCs have been engineered to express interferon-a (IFN-a) from an exogenous IFN-a coding sequence that is operably linked to both positive and negative expression control elements, wherein the control positive and negative control elements together restrict expression of the exogenous IFN-a coding sequence to differentiated myeloid progeny cells within a tumor microenvironment (TME), wherein the therapeutically effective number of autologous CD34+HSPCs is at least 1.5 E6 cells per kilo patient body weight (cells / kg).

[0129] In some embodiments, the positive expression control element comprises a Tie2 gene enhancer and promoter operably linked to the exogenous IFN-a coding sequence.

[0130] In some embodiments, the negative expression control element comprises at least one miRNA target sequence (mirT) engineered into the expressed IFN-a transcript. In some embodiments, each of the at least one mirT is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells.

[0131] In some embodiments, the negative expression control element comprises a plurality of mirTs engineered into the expressed IFN-a transcript. In some embodiments, each of the plurality of mirTs is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells.

[0132] In some embodiments, the plurality of mirTs engineered into the expressed IFN-a transcript has between one mirT to 10 mirTs. In some embodiments, the plurality of mirTs has at least one mirT, at least two mirTs, at least three mirTs, at least four mirTs, at least five mirTs, at least six mirTs, at least seven mirTs, at least eight mirTs, at least nine mirTs, or at least ten mirTs. In some embodiments, the plurality of mirTs engineered into the expressedIFN-a transcript has at least two mirTs. In some embodiments, the plurality of mirTs engineered into the expressed IFN-a transcript has one mirT, two mirTs, three mirTs, four mirTs, five mirTs, six mirTs, seven mirTs, eight mirTs, nine mirTs, or ten mirTs. In some embodiments, the plurality of mirTs engineered into the expressed IFN-a transcript has two mirTs. In some embodiments, the plurality of mirTs engineered into the expressed IFN-a transcript are located in different regions of the expressed IFN-a transcript.

[0133] In some embodiments, the negative expression control element has at least one mirT engineered into the expressed IFN-a transcript. In some embodiments, the at least one mirT is a mirT-126 target (miR-126T). In some embodiments, the at least one mirT is a miR-130a target (miR-130aT). In some embodiments, the at least one mirT is a mir-126T and a miR- 130a.

[0134] In some embodiments, the negative expression control has at least two mirT engineered into the expressed IFN-a transcript. In some embodiments, the negative control element has at least two mir-126T engineered into the expressed IFN-a transcript. In some embodiments, the negative control element has two mir-126T engineered into the expressed IFN-a transcript. In some embodiments, the negative expression control element has at least two mir-130aT engineered into the expressed IFN-a transcript. In some embodiments, the negative expression control element has two mir-130aT engineered into the expressed IFN-a transcript.

[0135] In some embodiments, the exogenous IFN-a coding sequence is selected from an IFN- al coding sequence, an IFN-a2 coding sequence, an IFN-a4 coding sequence, an IFN-a5 coding sequence, an IFN-a6 coding sequence, an IFN-a7 coding sequence, an IFN-a8 coding sequence, an IFN-alO coding sequence, an IFN-al3 coding sequence, an IFN-al4 coding sequence, an IFN-al6 coding sequence, an IFN-al7 coding sequence, and an IFN-a21 coding sequence. In some embodiments, the exogenous IFN-a coding sequence is an IFN-a2 coding sequence.

[0136] In some embodiments, the HSPCs have been engineered via integration of a lentivirus expression vector. In some embodiments, the lentivirus expression vector is a replicationdefective 3rd-generation pseudotyped vector. In some embodiments, the lentivirus is TIA126-LV.

[0137] In some embodiments, each of the mirT targets is encoded by a sequence in the integrated lentivirus expression vector that is transcribed into the IFN-a mRNA.

[0138] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is at least 1.5 E6 cells / kg. In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is 1.5 E6 cells / kg.

[0139] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is at least 2.0 E6 cells / kg. In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is 2.0 E6 cells / kg.

[0140] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is at least 2.5 E6 cells / kg. In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is 2.5 E6 cells / kg.

[0141] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is at least 3.0 E6 cells / kg. In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is 3.0 E6 cells / kg.

[0142] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is at least 3.5 E6 cells / kg. In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is 3.5 E6 cells / kg.

[0143] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is at least 4.0 E6 cells / kg. In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is 4.0 E6 cells / kg.

[0144] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs is at least 4.4 E6 cells / kg. In some embodiments, thetherapeutically effective number of the engineered autologous CD34+HSPCs is 4.4 E6 cells / kg.

[0145] In some embodiments, the therapeutically effective number of the engineered autologous CD34+HSPCs are intravenously administered to the patient.

[0146] In some embodiments, the method further comprises administering untransduced CD34+support cells in combination with the engineered CD34+HSPCs. In some embodiments, the untransduced CD34+support cells are HSPCs. In some embodiments, the untransduced CD34+HSPCs are autologous to the recipient.

[0147] In some embodiments, the untransduced CD34+support cells (e.g. supporter cells) and the autologous CD34+HSPCs to be transduced with the lentivirus expression vector are collected from the patient at the same time. In some embodiments, the untransduced CD34+support cells are autologous CD34+HSPCs from the patient. In some embodiments, the collected HSPCs from the patient are divided into two groups. In some embodiments, the two groups are (1) the untransduced CD34+support cells and (2) the autologous CD34+HSPCs to be transduced with the lentivirus expression vector. In some embodiments, the HSPCs are collected via leukapheresis.

[0148] In some embodiments, at least 1.5 E6 supporter cells / kg patient body weight are administered to the patient. In some embodiments, 1.5 E6 supporter cells / kg body weight are administered to the patient.

[0149] In some embodiments, at least 2.0 E6 supporter cells / kg patient body weight are administered to the patient. In some embodiments, 2.0 E6 supporter cells / kg patient body weight are administered to the patient.

[0150] In some embodiments, at least 2.5 E6 supporter cells / kg patient body weight are administered to the patient. In some embodiments, 2.5 E6 supporter cells / kg patient body weight are administered to the patient.

[0151] In some embodiments, at least 3.0 E6 supporter cells / kg patient body weight are administered to the patient. In some embodiments, 3.0 E6 supporter cells / kg patient body weight are administered to the patient.

[0152] In some embodiments, at least 3.5 E6 supporter cells / kg patient body weight are administered to the patient. In some embodiments, 3.5 E6 supporter cells / kg patient body weight are administered to the patient.

[0153] In some embodiments, at least 4.0 E6 supporter cells / kg patient body weight are administered to the patient. In some embodiments, 4.0 E6 supporter cells / kg patient body weight are administered to the patient.

[0154] In some embodiments, the untransduced CD34+support cells are administered intravenously to the patient.

[0155] In some embodiments, the untransduced CD34+support cells are administered to the patient prior to the administration of the engineered autologous CD34+HSPCs. In some embodiments, the untransduced CD34+support cells are administered to the patient at least two hours prior to the administration of the engineered autologous CD34+HSPCs. In some embodiments, the untransduced CD34+support cells are administered to the patient at least three hours prior to the administration of the engineered autologous CD34+HSPCs. In some embodiments, the untransduced CD34+support cells are administered to the patient at least four hours prior to the administration of the engineered autologous CD34+HSPCs. In some embodiments, the untransdcued CD34+support cells are administered to the patient at least five hours prior to the administration of the engineered autologous CD34+HSPCs. In some embodiments, the untransduced CD34+support cells are administered to the patient at least six hours prior to the administration of the engineered autologous CD34+HSPCs.

[0156] In some embodiments, the method further comprises the step, preceding administration of the engineered autologous CD34+cells, of administering at least one sub- myeloablative conditioning regimen. In some embodiments, the at least one sub- myeloablative conditions regimen is administered at least 6 days prior to the administration of the engineered autologous CD34+cells. In some embodiments, the at least one sub- myeloablative conditioning regimen is administered at least 5 days prior to the administration of the engineered autologous CD34+cells. In some embodiments, the at least one sub- myeloablative conditioning regimen is administrated at least 4 days prior to the administration of the engineered autologous CD34+cells. In some embodiments, the at least one sub-myeloablative conditioning regimen is administrated at least 3 days prior to the administration of the engineered autologous CD34+cells.

[0157] In some embodiments, the at least one sub-myeloablative conditioning regimen is selected from BCNU and thiotepa, busulfan and thiotepa, and busulfan without thiotepa. In some embodiments, the at least one sub-myeloablative conditioning regimen is BCNU and thiotepa. In some embodiments, the at least one sub-myeloablative conditioning regimen is busulfan and thiotepa. In some embodiments, the at least one sub-myeloablative conditioning regimen is bulfsan without thiotepa (e.g., bulsfan alone). In some embodiments, a single sub- myeloablative conditioning regimen is administered. In some embodiments, the single sub- myeloablative conditioning regimen is busulfan alone.

[0158] In some embodiments, the at least one sub-myeloablative condition regimen is administered to the patient intravenously, intramuscularly, orally, or intratum orally. In some embodiments, the at least one sub-myeloablative condition regimen is administered to the patient intravenously. In some embodiments, the at least one sub-myeloablative conditioning regimen is administered to the patient orally. In some embodiments, the single sub- myeloablative conditioning regimen is administered intravenously to the patient. In some embodiments, the single sub-myeloablative conditioning regimen is administered orally to the patient.

[0159] In some embodiments, the GBM is GBM with unmethylated MGMT promoter (uMGMT).

[0160] In some embodiments, the method further comprises the step, prior to administering the engineered autologous CD34+HSPCs, of surgically resecting the tumor mass of the patient.

[0161] In some embodiments, the method further comprises the step, before or after surgical resection and before administering the engineered autologous CD34+HSPCs, of administering radiotherapy to the patient. In some embodiments, the method further comprises the step, before surgical resection and before administering the engineered autologous CD34+HSPCs, of administering radiotherapy to the patient. In some embodiments, the method further comprises the step, after surgical resection and before administering the engineered autologous CD34+ HSPCs, of administering radiotherapy to the patient. In some embodiments, the method further comprises the step, before and after surgical resection and before administering the engineered autologous CD34+HSPCs, of administering radiotherapy to the patient.

[0162] In some embodiments, the radiotherapy is selected from external beam radiation therapy, stereotactic radiosurgery, brachytherapy, carbon ion radiotherapy, and laser interstitial thermal therapy. In some embodiments, the radiotherapy is external beam radiation therapy. In some embodiments, the external beam radiation therapy is selected from 3D conformal radiation therapy, intensity-modulated radiation therapy, image-guided radiation therapy, and proton beam therapy.

[0163] In some embodiments, the radiotherapy is not administered with a chemotherapy drug. In some embodiments, the radiotherapy is not administered with temozolomide.

[0164] In some embodiments, the method further comprises the step, subsequent to administering the CD34+engineered HSPCs, of surgically resecting the tumor mass of the patient.

[0165] In some embodiments, the administration of the therapeutically effective number of the engineered autologous CD34+HSPCs decreases the size of one or more GBM tumors in the patient. In some embodiments, administering the therapeutically effective number of the engineered autologous CD34+HSPCs decreases the size of the one or more GBM tumors in the patient by at least 5% compared to the size of the one or more GBM tumors in the patient prior to administering the therapeutically effective number of the engineered autologous CD34+HSPCs. In some embodiments, administering the therapeutically effective number of the engineered autologous CD34+HSPCs decreases the size of the one or more GBM tumors in the patient by at least 5%, at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, or at least 100% compared to the size of the one or more GBM tumors in the patient prior to the administration of the therapeutically effective number of the engineered autologous CD34+HSPCs.5.2. Examples

[0166] Below are examples of specific embodiments for carrying out the present invention. The examples are offered for illustrative purposes only and are not intended to limit the scope of the present invention in any way. The practice of the present invention will employ, unless otherwise indicated, conventional methods of molecular biology, cell biology, biochemistry,genetics, cancer biology, and pharmacology, within the skill of the art. Such techniques are explained fully in the literature.5.2.1. A Phase I / IIa Dose Escalation Study Evaluating the Safety and Efficacy of Autologous CD34+-Enriched HSPCs Genetically Modified With Human lnterferon-«2 in Patients With Glioblastoma Multiforme and Unmethylated MGMT Gene Promoter (NCT03866109)

[0167] A non-randomized, open label, multicenter, phase I / IIa, prospective study was conducted involving a single infusion of “Temferon™” - autologous CD34+-enriched HSPCs transduced with a third generation lentiviral vector driving myeloid-specific IFN-a2 expression - to patients affected by GBM who have an unmethylated MGMT promoter (uMGMT). The clinicaltrials.gov designation for the trial is NCT03866109.5.2.1.1 Protocol Synopsis5.2.1.2 Protocol outline

[0168] Potentially eligible patients are being identified after surgical resection of GBM once the MGMT promoter methylator status is known. Once written informed consent is obtained and screening procedures have been completed, harvesting of HSPCs is performed. A standard of care regimen lasting approximately 6 weeks is then administered. During this time, Temferon™ manufacturing occurs. Following completion of radiotherapy, patients are admitted for receipt of a conditioning regimen consisting of BCNU and thiotepa (Cohorts 1- 6), busulfan and thiotepa (Cohort 5), or busulfan (Cohorts 7 and 8). This is followed by administration of Temferon™ In- patient monitoring is done until hematological recovery occurs. Thereafter, regular follow-up of patients occurs up to 2 years (+720 days) with the majority of assessments and procedures. At the+720 day visit, patients will be invited to participate in a long term follow-up study which will last for an additional 6 years.

[0169] Description and the current disposition of the trial cohorts is illustrated in FIG. 2.5.2.1.3 Temferon™5.2.1.3.1 TIA126 Lentiviral Vector

[0170] The TIA126-LV is a replication-defective 3rd-generation pseudotyped vector. It consists of a core of human immunodeficiency virus (HIV-1) structural proteins and enzymes. Specifically, it contains the envelope of the vesicular stomatitis virus-G (VSV-G), and the HIV-1 genome consisting in the cis-acting sequences without viral genes and one expression cassette for the IFN-a transgene. The three vector components (core, envelope and genome) are transiently expressed in vector producer cells by four different constructs: two core packaging constructs, the envelope construct and the transfer vector construct. The transfer vector construct encodes the human IFN-a2 gene under the control of the HSA.pTEK / s promoter and hsa.Tie2.e / s enhancer belonging to the human Tie2 locus (TEK) that encodes for the Tie2 angiopoietin receptor. The introduction of TEK promoter and enhancer allows the selective expression of human IFN-a2 by bone marrow (BM)-derivedmyeloid cells characterized by the Tie2 promoter activation. Moreover, the expression of hIFN-a2 gene is controlled by a post-transcriptional regulation layer represented by two target sequences of miR-126, which prevents expression in HSPCs where the Tie2 enhancer / promoter may have some activity.TIA126-LV Manufacturing Process

[0171] The TIA126-LV vector is produced starting from a packaging cell line transiently transfected with 4 plasmids encoding for two core packaging constructs (pKLGag / pol and pKRev), the envelope construct (pK.G) and the transfer vector construct (p.IFNa). A series of cell supernatant harvests, filtration and downstream purification occurs with the aim of degrading plasmid DNA, host cell derived DNA, host cell proteins, serum-derived proteins and other potential contaminants.

[0172] The resulting LV preparation undergoes two subsequent, sterilizing, 0.2 pm filtrations and aseptic filling. The purified vector preparation is stored at < -65°C.Transfer vector construct (p.IFN-a)

[0173] The transfer vector construct (p.IFN-a plasmid) encodes for the therapeutic IFN-a complementary DNA (cDNA) and for the sequences necessary for efficient encapsidation, reverse transcription, integration of the vector genome, and transgene expression. The lentiviral vector genome is designed with a self-inactivating mechanism (SIN) guaranteed by a deletion of 400 bp in the U3 region of the 3' long terminal repeat (LTR). The introduction of this deletion abolishes the production of full-length RNA vector in transduced cells by the reverse transcription process, thus minimizing the risk of RCL emergence and insertional mutagenesis. Moreover, the SIN design improves vector performance as the transgene expression control is exclusively dependent on the promoter introduced into the vector, and significantly alleviates concerns for transcriptional activation of genes flanking the integration site.

[0174] Human IFN-a expression is under the control of the human Tie2 enhancer / promoter and of a post- transcriptional regulation layer constituted by two miRNA 126 target sequences. In order to enhance transgene expression, a mutated Woodchuck hepatitis virus post-transcriptional regulatory element (WPRE) has been inserted downstream of the hlFN- a2 cDNA. The mutated form of the WPRE was chosen to eliminate the risk of potentialoncogenic activity. The transcription of the full-length vector messenger RNA (mRNA) in transfected vector-producer cells is driven by the pCCL hybrid promoter, which contains the cytomegalovirus (CMV) enhancer and promoter sequences upstream to the LTR R (U5) region of HIV-1. The construct also contains, from 5' to 3', the leader sequence, SD site, extended packaging sequence, the Rev Responsive Element (RRE) element, the SA site taken from the tat and rev second exon and the cPPT-CTS (Central Polypurine Tract- sequence involved in the nuclear transport of the viral genome) (Follenzi et al., 2000). A portion of the gag gene was retained as part of the extended packaging signal for optimal transduction.

[0175] A circular map of the p.IFNa plasmid is presented in FIG. 1.5.2.1.3.2 Temferon™ drug substance

[0176] The Drug Substance (DS) is defined as autologous CD34+HSPCs genetically modified with the TIA126-LV encoding for the human IFN-a2 gene. The DS is prepared ex vivo, by exposing CD34+HSPCs taken from the patient's peripheral blood or bone marrow to transduction with the TIA126-LV. The manufacturing process of Temferon™ DS can be divided in two steps: (1) CD34+cell selection and (2) stimulation of autologous CD34+cells, transduction and final DS preparation.5.2.1.3.3 Temferon™ drug product

[0177] The Drug Product, Temferon™, consists of frozen autologous CD34+hematopoietic stem and progenitor cells (HSPCs) genetically modified with the lentiviral vector TIA126- LV, encoding for the human interferon- a2 (IFN-a2) gene.5.2.1.4 Detailed Clinical Trial Design, Patient and Product Characteristics, Feasibility

[0178] A clinical phase I / IIa dose escalation study was initiated to evaluate the safety and efficacy of autological CD34+-enriched HSPCs transduced with a third-generation lentiviral vector driving myeloid-specific IFN-a2 expression (“Temferon™”) in newly diagnosed patients with GBM with unmethylated MGMT promoter (MGMTunmeth), representing the worst prognostic condition. The starting Temferon™ dose was set to reflect the minimum graft chimerism associated with some anti-tumor effect in mouse studies, and a dose escalation was planned until a chimerism of about 50% was reached. Inclusion and exclusion criteria are listed in Table 2.Table 2. Inclusion / Exclusion Criteria

[0179] In short, patients had surgically resectable disease, no contraindications for adjuvant radiotherapy, and good general health with a life expectancy of at least 6 months. Patients were enrolled after histopathologic confirmation upon primary surgical resection and, following informed consent, underwent HSPC mobilization and collection during 1 or 2 leukapheresis sessions (FIG. 10A). Part of the collection (> 5xl06CD34+ / kg) was cryopreserved as unfractionated apheresis, while the rest was used for Temferon™ manufacturing. The transduction process was designed to achieve VCN around 1 (range 0.3- 1.6) after a short culture time in medium widely used in prior HSC gene therapy trials.Patients, tumors and products characteristics are summarized in Tables 3-5, respectively. All patients were of European origin.Table 3. Baseline patient characteristics.Table 4. Tumor characteristics at diagnosis.Table 5. Product Characteristics2Two consecutive days of apheresis to obtain the target dose*: Total Temferon1 2dose was below the protocol-defined target, due to poor mobilization and below average process yield1': Total Temferon™ dose was increased based on low VCN, as defined in the protocol

[0180] While the drug product underwent release testing, patients received whole brain radiotherapy (60 Gray (Gy) over 6 weeks) omitting Temozolomide considering that itsbenefit in MGMTunmethGBM is expected to be none to minimal. Temferon™ (0.5xl06cells / kg to 4xl06cells / kg) was administered intravenously following sub-myeloablative conditioning chemotherapy, and at least 3 hours after infusion of a fixed dose of unfractionated apheresis (>2xl06to 3xl06CD34+ / kg), sufficient to allow hematologic recovery. The study enrolled 8 cohorts of 3 patients (FIG. 10B). Cohorts 1 to 4 and 6 used Carmustin (BCNU) and Thiotepa as conditioning chemotherapy, whereby the target dose of Temferon™ was escalated from 0.5xl05cells / kg to 3xl06cells / kg. The initial choice of conditioning regimen was based on its established clinical use in cerebral lymphoma, ensuring drug access to the CNS and potential effects on stimulating macrophage turnover in the GBM TME. Cohort 5 received the same cell doses as cohort 4, but using Busulfan, instead of BCNU, in combination with Thiotepa, to explore whether sub-myeloablative doses of Busulfan positively affected transduced HSC engraftment. In cohorts 7 and 8, busulfan monotherapy was used for conditioning, to reduce overall side effects, and a further dose escalation of Temferon™ from 3xl06cells / kg to 4xl06cells / kg was tested. Patients received a total CD34+cell dose (sum of Temferon™ and supporter cells) of approximately 4xl06cells / kg to 8xl06cells / kg, with a median proportion of TIA126-transduced CD34+cells that ranged between 8% (cohort 1) and 32% (cohort 8), as estimated from the cell counts at infusion and the proportion of transduced colony-forming units of the drug product (Table 5).

[0181] Twenty-four out of 34 screened patients (70.6%) were infused with Temferon™. Reasons for drop-out before infusion were screening failure (5 patients), withdrawal of consent (2 patients), sudden death likely due to cardiac arrest during radiotherapy, (1 patient), and ineligibility for conditioning chemotherapy due to poor performance status associated with disease progression (two patients).5.2.1.5 Engraftment in Patients and Clinical Outcome

[0182] All patients infused with Temferon™ showed the presence of gene-marked cells, e.g., Temferon™ engraftment, in the blood and bone marrow, with most of the transduced, engrafted clonogenic progenitors carrying a single integrant of TIA126, as expected from the design of the ex vivo transduction protocol (FIG. 10C, FIG. 10D and FIG. 11A-11E). Engraftment of TIA126-transduced cells, as estimated from the VCN in CD33+bone marrow progenitors, was highest at day +30 and generally well corresponded to the expected input of TIA126-transduced CD34+cells, with some variability between individual patients, indicating no obvious competitive disadvantage of the transduced vs. unmanipulatedsupporter cells (FIG. IOC). TIA-126+ cells then declined at subsequent timepoints and stabilized remaining detectable until the last follow-up in most patients in cohorts 3-8. Since Temferon™ cells in the drug product were homogenously transduced at VCN: 1, the initial decline between 1 and 3 months is likely due to the choice of sub-myeloablative conditioning, in which endogenous HSPCs that have survived chemotherapy recover and progressively dilute the output of the TIA126-transduced graft. In vivo gene marking strongly increased from dose level 1 to 3 and then plateaued until cohorts 7 and 8, where some further increase, albeit highly variable among patients, was observed, mostly at longer follow-up times. Note that cohorts 3 to 5 received the same amount of Temferon™ with different conditioning regimens or proportion of supporter cells. Both BCNU and Busulfan, with or without Thiotepa, enabled long-term engraftment of genetically engineered progenitor cells. To understand whether there were differences in graft stability between the conditioning regimens tested, the relative VCN decrease in granulocytes as a short-lived surrogate for bone marrow progenitors was assessed (FIG. HE). Patients from cohorts 7 and 8 (Busulfan monotherapy) showed a trend for more stable engraftment of genetically engineered cells, at least until month 3 post infusion.

[0183] Despite significant in vivo gene marking with TIA126, reaching up to 40-50% of blood leukocytes in some patients of the last cohorts (e.g., one patient in cohort 7), plasma IFN-a levels remained low, ranging up to 30pg / mL within the first 3 months and without obvious dependence on the engrafted VCN (FIG. 10D, FIG. 10E). An increase in interferonresponsive gene signatures (IRGs) was seen in blood and bone marrow leukocytes after Temferon™ infusion, which remained relatively stable over time and did not differ significantly between cohorts (FIG. 12A-12E). The induction of IFN signaling within the most IFN-sensitive hematopoietic lineages likely provided an indication of biological activity of the transgene, which is, however, subjected to tight regulation by the TIA126 vector that limits clinically-significant systemic exposure to IFN-a.

[0184] At the data cut-off, median elapsed time since Temferon™ infusion was 39 months, which was over 15 months in 19 of the 24 dosed patients. Only cohort 8 had less than 1 year of follow-up. Median overall and progression-free survival was 16.8 and 9.5 months, respectively, from the primary surgical procedure at which the diagnosis was made (FIG.10F), and 13.2 and 6.2 months from Temferon™ infusion. Eight out of 19 evaluable patients (42%) survived beyond 15 months from the Temferon™ infusion, with 5 patientsreaching over 2 years (Fig.lG) and 1 patient 3 years. These preliminary findings on a small cohort of patients are encouraging, as they are significantly better than observed for a much larger cohort of MGMTunmethGBM patients historically treated in the same hospital by standard-of-care.5.2.1.6 Safety and Tolerability

[0185] Temferon™ infusion was well tolerated, and there were no adverse events associated with the product in the 24 hours following the infusion (FIG. 13). Importantly, patients did not experience cytokine-related side effects such as flu-like symptoms, pyrexia, nausea, and anorexia, which are frequently observed following the administration of recombinant IFN-a. Hematologic recovery after Temferon™ administration was achieved by day +30 in 100% of the patients and was consistent with autologous stem cell transplantation. Spontaneous neutrophil and platelet recovery occurred on median day +14 and +19 respectively, with no significant differences across all administered Temferon™ dose levels (FIG. 14A, 14B). Notably, in patients treated with busulfan alone (cohorts 7 and 8), there was a delayed decline in neutrophil counts, resulting in a shortened duration of grade 4 neutropenia. In addition, the total lymphocyte count remained stable in these cohorts (FIG. 15A) in line with previous observations that busulfan does not cause acute lymphodepletion. Patients received on average 5 platelet transfusions to maintain platelet counts above 50xl09 / L throughout the treatment phase, a prudent threshold chosen to prevent intracranial bleedings in patients with active CNS malignancy (FIG. 14B). In the long term, blood counts and hemoglobin levels stabilized in the lower normal range, with no difference between patients receiving different Temferon™ doses (FIG. 15B). T helper cell (CD4+) counts took longest to recover and were depleted to a similar extent in all cohorts (FIG. 15C).

[0186] Treatment-emergent adverse events (TEAE) were reported in all patients (FIG. 16), the majority of which (75%) were mild or moderate (CTCAEv.5 grade 1 or 2) and considered unrelated to Temferon™ (95%), while half of all events were related to conditioning chemotherapy. Events were similarly distributed across all cohorts, with no association to Temferon™ dose. However, patients that received single alkylator conditioning with Busulfan, had fewer TEAEs, notably less infections. A total of 32 serious adverse events (SAE) were reported in 16 out of 24 patients, mainly infections and cytopenias occurring before day +90, and neurologic and general disorders thereafter, because of GBM disease progression (FIG. 14C). Three patients, all of which received double alkylator conditioning,died within 4 months from Temferon™ administration from infectious complications, corresponding to a transplant-related mortality of 12.5% in this patient population. Another patient experienced sudden death on day +402 post Temferon™ infusion, likely related to disease progression. One SAE, namely a persistent elevation of liver enzymes >10 fold above the upper limit of normal, was judged as possibly related to Temferon™. No autoimmune manifestations consequent to Temferon™ administration were recorded. No dose limiting toxicities were observed with Temferon™ doses up to 4.0xl06cells / kg. It is noteworthy that patients who underwent second line chemotherapy or second craniotomy for GBM progression had a normal clinical course. This confirms that treatment with Temferon™ does not appear to have adverse effects on marrow regeneration or wound healing.

[0187] Vector integration site analysis on longitudinal blood and BM samples showed a genomic integration profile typical of lentiviral vectors, without any evidence for insertional mutagenesis or clonal dominance (FIG. 17A-17C).5.2.1.7 Temferon™ Progeny Enter the CNS and Induce IFN-aExposure within the Tumor

[0188] To evaluate the penetration and activity of genetically engineered cells within the CNS, IFN-a levels in the cerebrospinal fluid were measured over time (FIG. 14D). IFN-a was undetectable (<1.95 pg / ml) at screening in all but one patient but was measurable in 6 / 12 (50%) evaluable patients on day +30, in 11 / 15 (73%) on day +90, and in 100% of patients on day +180 and day +360 (9 / 9 and 3 / 3 respectively). Peak levels were often observed on day +180, consistent with increasing recruitment of genetically engineered hematopoietic cells into the TME over at least 6 months. Tissue samples were available from 8 patients after Temferon™ infusion, 1 of which was from an autopsy and 7 from patients who underwent second surgery as part of their next line of treatment (FIG. 14E). DNA extracted from the tumors (either formalin-fixed, paraffin-embedded tissue sections, and / or purified CD45+cells from dissociated tumors) showed the presence of vector copies in all samples (VCN range: 0.01-0.06), except for one patient from cohort 1, which had the lowest transduced cell engraftment in the study (FIG. 14F). Considering that most engrafted transduced cells had a single vector integrant (FIG. 11D), this meant that up to about 6% of the tumor consisted of transduced cells. To confirm the presence of TEMs in GBM, flow cytometry was performed on CD45+tumor-infiltrating leukocytes from second surgery samples from patients enrolled in this trial or from other GBM patients undergoing standard-of-care treatment (FIG. 18A-18D) Surface expression of TIE2 was detectable on a variable proportion of TAMs, ranging from 2% to 25% (FIG. 14F and FIG. 18C, 18D). To investigate exposure to IFN-a in the TME, the expression of three representative IRGs, namely 2'-5'-Oligoadenylate Synthetase 1 (OAS1), Interferon Induced Protein With Tetratricopeptide Repeats 1 (IFIT1) and Interferon Regulatory Factor 7 (IRF7), were measured in RNA extracted from formalin-fixed, paraffin- embedded tissue sections or snap-frozen tumors (FIG. 14F and FIG. 19A, 19B). Tumor biopsies from second surgeries showed up to 10-fold higher expression of IRGs compared to patient-matched tumor samples from primary resection. Of note, one patient, which comprised the lowest fraction of TEMs, showed the least IRG induction. These data suggested IFN-a release by transduced, TIE2-expressing macrophages (TEMs) within the TME.

[0189] Histopathologic analysis was performed on tumor samples from first and second surgery of trial patients, as well as recurrent GBM from patients treated with standard of care (Table 6) and undergoing second surgery (“internal dataset”, FIG. 20A-20E). As compared to recurrences in the SoC patients, progressing lesions of Temferon™ treated patients showed increased extent of necrosis not attributable to prior radiotherapy and substantially lower angiogenesis.

[0190] Taken together, these data indicate that Temferon™ progeny are recruited to the GBM TME, where there was evidence of induced local exposure to IFN-a and its downstream effects, conceivably due to activation of the vector TIE2 enhancer / promoter in at least a subset of TAMs.5.2.1.8 Temferon™ Reprograms the Tumor Microenvironment

[0191] Next, single cell RNA sequencing (scRNAseq) was performed on dissociated tumors from n=6 trial patients after Temferon™ infusion, as well as n=6 recurrent GBM from patients treated with standard of care (Table 6) and undergoing second surgery (“internal dataset”, FIG. 20A-20E).Table 6. Characteristics of Recurrence Samples from Control Patients*M: methylated MGMT promoter; NM: non-methylated MGMT promoter

[0192] Cells were sorted for CD45 expression to enrich for hematopoietic stromal cells (CD45+) vs. tumor cells (CD45neg), respectively. To obtain a more comprehensive analysis, additional public scRNAseq datasets from GBM patients were integrated to create a GBM atlas with a simplified, manually curated classification of cell populations (FIG. 21A-21H, FIG. 22A-22E)

[0193] The CD45negcompartment that mostly comprised heterogenous tumor cells was analyzed first, and the GBM atlas showed 4 macro-clusters based on the predominant tumor cell state: G-I, oligodendrocytic; G-II, astrocytic; G-III, neural; G-IV, mesenchymal-like (FIG. 22A-22C). Except for sample TEM-G13 (cluster G-I), the other samples from the trial were assigned to G-II, the largest macro-cluster in the GBM atlas with broad representation of different tumor cell states.

[0194] The analysis then specifically focused on myeloid and T lymphoid cells, respectively. Myeloid cells mostly comprised macrophages, which distributed in an almost continuous landscape. The following major macrophage subpopulations were distinguished: MIF+ and APOC+ macrophages, which expressed several other genes associated with immunosuppressive functions, such as SPP1 and TREM2; Inflammatory and intermediate early response (IER) gene macrophages, characterized by the expression of activation markers (JUN and FOS) and pro-inflammatory cytokines and signals (IL1B, CCL4, NFKB1 A); SELENOP+ macrophages, that showed an intermediate phenotype, between the above mentioned, more polarized subsets (FIG. 21A-21B). More globally, the myeloid landscape in the GBM atlas highlighted 4 macro-clusters: M-I, monocyte-rich, mostly seen in recurrent GBM; M-II, MIF+ macrophage-rich; M-III, microglia / APOC+ macrophage-rich; M-IV, inflammatory macrophage-rich (FIG. 21C). Five out of 6 patients that received Temferon™ provided GBM samples that clustered in M-IV, while only 1 patient (TEM-G20) clustered in M-II. A second sample from a progressing lesion from patient TEM-G11 also clustered in M-II (progressing lesion 1, see below). In contrast, only 34% of control cases clustered as M-IV, with most categorized as M-II or M-III. Strikingly, the fraction ofinflammatory macrophages within the myeloid compartment was significantly increased in GBM of patients that received Temferon™ as compared to all other samples, accompanied by a proportionally decreased fraction of MIF+ and APOC+ macrophages (FIG. 21D; FIG. 22D)

[0195] The tumor infiltrating T lymphocyte (TIL) compartment of GBM was generally poor and could only be obtained from 38 patients in the atlas, mostly from recurrent tumors (FIG. 21E-21F). It mostly contained naive and CD4+cells, as well as granzyme K (GZMK)+ CD8 effector cells. Strikingly, 5 out of 7 Temferon™ GBM samples formed a nearly private macro-cluster (T-I), which differed from the most prevalent T-II cluster by an increased representation of GZMK+ effector / exhausted CD8 T lymphocytes (FIG. 21G). Indeed, T cell composition was different in patients who received Temferon™ with a significant increase in GZMK+ effector / exhausted, naive / activated and y8T-NK cells, and a proportionate decrease mostly in CD4 T cells and ZNF683 effector CD8 T cells (FIG. 21H, FIG. 22E).

[0196] To further characterize the Temferon™ TIL compartment, single cell TCR sequencing was performed on the internal dataset. After transferring cell annotations defined in the GBM atlas onto the internal dataset (FIG. 12A), the CD4 compartment was observed to generally contain small-medium sized clones, whereas large and hyperexpanded clones were mostly found within the CD8 effector compartment, particularly in the GZMK+ effector / exhausted CD8 subset, which was the most represented in Temferon™ samples (FIG. 12B) Several expanded clones were selectively enriched inside the tumor, as determined by bulk TCR sequencing (clonoSEQ) of longitudinal tumor and blood samples, and either appeared de novo in the recurring lesion or were already present in the primary lesion and could be tracked in the blood during the follow-up (FIG. 23).

[0197] Intra-cluster gene set enrichment analysis between Temferon™-treated and control tumors showed universal enrichment of interferon- and inflammatory response genes, together with associated signaling pathways across most subpopulations, including the immunosuppressive MIF+ and APOC+ macrophages, confirming pervasive IFN-a exposure within the entire tumor ecosystem (FIG. 12D). Moreover, Temferon™ treated cells showed down-regulation of oxidative phosphorylation and upregulation of hypoxia and apoptosis across tumor and several stromal cell populations, with MYC targets also reduced, suggesting anti-angiogenic, cytotoxic and anti-proliferative effects.

[0198] To further assess distinct ligand-receptor signalling that mediates tumor immune microenvironment reprogramming was applied in the context of Temferon™ treatment, NicheNet, a method that predicted which ligands of the sender cells (myeloid cells) were most likely to potentially affect the expression of target genes (DEGs upregulated in Temferon™ samples vs Controls) in the receiver cells (CD8+GZMK+Eff / Exh T cells). Topranked up-regulated ligand / receptor pairs (FIG. 12E) were involved in IFN-a signaling (IFITM1-CD81), TNF-signalling (GNR-TNFRSF1B), antigen-presentation (HLA.E- CD8A / KLRK1, HLADR.B5-LAG3, HLA.A-CD8A, B2M-TAP1) and T-cell activation (CD55-ADGRE5). Of note, this analysis highlighted the CXCL16 / CXCR6 axis, previously correlated with T cell infiltration and the generation of antitumor immunity in glioblastoma.

[0199] These data uncovered a specific impact of Temferon™ on the GBM macrophages and TIL compartments favoring inflammatory activation of the former and induction and accumulation of T cell subpopulations with several features of tumor reactive CD8 effector cells.5.2.1.9 Immune profiling results

[0200] Tumor tissue samples obtained via a second surgery from clinical trial subjects have been evaluated following treatment with Temferon™, at single cell resolution.

[0201] Single cell RNA sequencing (scRNA) analysis demonstrates that the tumor and tumor microenvironment of patients treated with Temferon™ sensed the local release of IFN-a, as evidenced by the induction of interferon-response genes, both in the tumor and in the TME, including tumor-associated myeloid cells and T-cells. FIG. 3 presents Gene Set Enrichment analysis performed on ranked log-fold change gene lists obtained from intra-population differential expression analysis between the Temferon™-treated group and subjects outside the clinical trial who received Standard-of-Care (SoC) treatment without Temferon™ administration, showing a widespread and strongly significant (NES > 0; p.adj(BH) < 0.01) interferon response and inflammatory response in most of the cell populations from both the tumor microenvironment (TME) and the glioblastoma (GBM) tumor cells themselves.

[0202] In the myeloid cell compartment, a reprogramming toward a pro-inflammatory (Ml- like) phenotype was observed in Temferon™ patients. FIG. 4A illustrates myeloid TME reprogramming to Ml -like populations from M2-like populations, with (a) a higherproportion of pro-inflammatory macrophages (M) in the Temferon™ group and (b) a higher proportion of APOC1_2-M in the standard-of-care treated (untreated) group.

[0203] FIG. 4B presents superimposed UMAP plots of scRNA seq data from post- Temferon™ treatment “second surgery” biopsies of two lesions from the same clinical trial subject, one of which had progressed since first surgery (“Progressed”) and one that had not progressed (“Stable”). The comparison shows a noticeable skewing toward an Ml-like phenotype in the Stable lesion as compared to the Progressed lesion, which was conversely enriched in hypoxic (M2 -like) macrophages.

[0204] In addition to the remodeling of the myeloid compartment, a consistent increase in CD8+lymphocytes with effector phenotype was observed in Temferon™ patients. FIG. 5A is a heat map showing levels of expression of various T cell markers (listed on the y-axis) in tumor samples drawn from individual subjects in the Temferon™ clinical trial (indicated by an upward arrow on the x-axis), and in tumor samples from non-trial subjects with primary, recurrent, or unknown stage tumor status who had not been treated with Temferon™. Tumors from the Temferon™-treated group showed an increased proportion of CD8+cells. In addition, a consistent increase in CD8+lymphocytes with effector phenotype was observed in Temferon™ patients as compared to SoC controls (FIGs. 5A and 5B), including an increase of effector and effector memory cells as compared to the Gbmap-core control dataset (FIG. 5B)

[0205] To identify expanded and putative tumor-reactive clones, TCR repertoire analysis of the TCR beta chain (TCRB) was added to the scRNAseq analysis to assess the presence of the NeoTCR8_ALL gene signature previously identified by Lowery etal., Science 2022 (FIG. 6A), and demonstrated that clonotype abundance increases with T cell differentiation towards a CD8 effector state (FIG. 6B). The fraction of putatively tumor-reactive cells that are associated with expanded clones is 2-4 fold enriched in Temferon™-treated patients compared to SoC-treated patients. Mapping of NeoTCR8_ALL signature (module score - default settings) also highlights that putative tumor reactive TILs (trTILs) are mainly CD8 Effector T cells (FIG. 6B). Tumor reactivity provided by the NeoTCR8_ALL signature matches with the region of maximal clonal expansion.

[0206] Preliminary evidence shows that Temferon™ administered at a dose of >2 x 106CD34+cells / kg favors T cell recruitment.

[0207] The expression of key immune checkpoint molecules such as PD-1, PD-L1 and CTLA4 was analyzed both in the tumor and in TME (myeloid and T cells) in Temferon™- treated and SoC-treated patients, using a discovery dataset of 5 Temferon™ patients and 6 SoC treated patients. No differences were observed between the Temferon™-treated and the SoC-treated group in bulk tumor samples or in the myeloid compartment for any of the evaluated molecules. However, as shown in FIGs. 7A, 7B, and 7C, when expression of these markers in T-cells was analyzed, a higher expression of PDCD1 (PD-1) was observed in CD8-Effector and CD8-IFN-a populations in Temferon™ patients. Interestingly, most of the PDCD1+(PD-1+) T-cells belonged to the stable lesion of the 2-lesions patient (Figure 7C). This finding could suggest that IFN-a may promote the generation of tumor-reactive T cells that are associated with a temporary response, with the trTILs upregulating PD1 expression as a consequence of continuous antigen stimulation.

[0208] Preliminary analyses on the evaluation of PD1+ and CTLA4+ populations showed that Temferon™-treated patients have a 2-fold enriched PD1+ vs. CTLA4+ median ratio compared to the SoC-treated group (FIG. 8). This could suggest that the Temferon™ group is more prone to post-tumor reactivity exhaustion, and provides a particular rationale for combining Temferon™ treatment with anti-PDl antibodies.

[0209] Thereafter, among the PD1+ and CTLA4+ CD8-T cells, tumor reactivity based on the Lowery gene signature and clonotype identity and frequency obtained from TCR repertoire analysis was analyzed, both at single-cell level (2nd surgery) and at bulk level (1stsurgery, 2nd surgery, and follow up timepoints including screening and baseline on peripheral blood samples). Preliminary results, at the clonotype level, suggest the absence of correlation between the median expression values of PD-1 or CTLA4 and the different classes of clonotypes expanded from 1st surgery to 2nd surgery across follow up gene therapy timepoints. However, when the stable lesion (the most enriched in PD1+ cells) was focused on, restricting the analysis to the tumor-reactive CD8+T clonotype that is absent in the peripheral blood (PB) samples to exclude putative by-stander tumor-infiltrating-lymphocytes (TILs), a strong enrichment of CD8+tumor reactive T-cells in the stable lesion was observed compared to progressed one. In particular, most of the tumor reactive TILs specific to 2nd surgery and not present in peripheral blood samples map into the CD8-Effector population, highly enriched in the stable lesion, whereas TILs located in the CD8-CTLs, most abundant in the progressed lesion, are putatively by-standers and shared with peripheral blood.

[0210] FIG. 9 is a UMAP plot of cells in the stable and progressed lesion that belong to clonotypes that are private to second surgery samples and not present in peripheral blood (PB associated samples). In light-blue are cells that are defined as expanded (that is, in a clonotype with frequency > 0.1%) tumor reactive TILs (trTILs) according to the Lowery et al. NeoTCR8_ALL signature. The quota of trTILs is much higher in the stable versus the progressed lesion.

[0211] Differential gene expression between tumor reactive TILs and bystander TILs within the CD8-Effector T cell population in the stable lesion show up-regulation of exhaustion molecules including PD1 and TOX, suggesting that these cells had therapeutically important prior engagement with antigen.

[0212] Taken together, these findings demonstrate that Ml-biased myeloid reprogramming in the tumor microenvironment by Temferon™ leads to activation of CD8+effector T cells and expansion of tumor-reactive T cell clones in the tumor. The upregulation of T cell PD-1 expression among these tumor-reactive T cells specifically identifies the PD-1 / PD-L1 axis as an attractive target for concurrent intervention, with coadministration of Temferon™ and inhibitors of PD-1 / PD-L1 interaction likely to lead to greater efficacy in GBM than either alone.

[0213] These findings also predict that co-administration of Temferon™ and checkpoint inhibitors will provide greater efficacy than either alone in the treatment of other classically immunologically “cold” solid tumors such as renal cell carcinoma, melanoma, high grade osteosarcoma, non-small cell lung cancer (NSCLC), breast cancers, squamous cell carcinoma of the head and neck, bladder cancer, liver and intrahepatic bile duct cancer, gastroesophageal adenocarcinoma / squamous cell cancer, mesothelioma, liver metastases from primary colorectal, breast and urothelial cancers, and melanoma, and epithelial ovarian cancer.

[0214] Moreover, it has previously been shown that a sizable fraction of patients with immunologically “hot” tumors do not respond to immune checkpoint inhibitors due to a nonfunctional T cell repertoire. In this regard, Temferon™ - by recruiting and activating newly engaged CD8 effector T cells and tumor-reactive T cells, which will then upregulate PD1 expression - will create a favorable substrate to further enhance the efficacy of immune checkpoint inhibition in such patients. These findings thus also have relevance for the treatment of immunological “hot” tumors, such as renal cell carcinoma, melanoma, non-smallcell lung cancer (NSCLC), squamous cell carcinoma of the head and neck and bladder cancer by co-administration of Temferon™ and checkpoint inhibitors.5.2.1.10 Evidence for Development of Anti-Tumor Immunity

[0215] 2 distinct tumor lesions were simultaneously analyzed that showed a different local response in patient TEM-G11, allowing for an intra-patient comparison of TME characteristics associated with stability vs. progression. One month after surgery, 4 distinct target lesions (TL) were detectable by MRI, with TL1 and TL2 being larger (FIG. 24A). Following radiotherapy and Temferon™ administration, TL2 and TL3 shrank and TL1 appeared more necrotic. During further follow up, TL1 and TL4 remained stable, while TL2 increased in volume from day +90 and continued to progress triggering the second surgical intervention on day +205 and eventually causing the death of the patient on day +340. The expression of IRGs was measured on snap-frozen biopsies from TL1 and TL2 using digital droplet PCR. In line with the qPCR results (FIG. 21C), the progressing TL2 showed low expression (FIG. 24B). On the other hand, the stable TL1 showed high expression of IRGs, suggesting differences in the TME between TL1 and TL2. Flow cytometry showed a lower proportion of TEMs in the progressing lesion as compared to the stable lesion (3% vs 16%, respectively), which could explain the low IFN-a signature despite the presence of transduced cells in this tumor (FIG. 24C). Strikingly, single cell RNA sequencing confirmed a different composition of the TAM compartment between these 2 TLs, with an inverted MIF+ to inflammatory macrophage ratio (FIG. 24D). Functional MRI analysis on day +120, when both TLs had a similar size, showed a lower standardized cerebral blood volume (sCB V) in the progressing TL2 suggestive of lower perfusion and more hypoxic TME (FIG. 24E). Indeed, MIF+ macrophages showed a strong hypoxic gene signature. The myeloid compartment of both lesions was interrogated with a TEM gene signature derived from bulk RNA sequencing of TIE2-positive versus -negative CD45+CD14+sorted cells from a control GBM sample (FIG. 18), which captured a broad representation of genes marking immunosuppressive TAMs (FIG. 25). Intriguingly, a reshaping of the distribution of signature positive cells was found, which were skewed toward inflammatory TAMs in the stable lesion (FIG. 26). The enrichment of the TIE2 signature in inflammatory TAMs in TL1 was consistent with genetic engineering of TEMs toward IFN-a release, which beside its paracrine effect in the TME, may reprogram these immunosuppressive macrophages when reaching a sufficient fraction of cells.

[0216] Regarding the T cell compartment, the stable TL1 also contained a higher proportion of neoTCR8 signature+, putatively tumor-reactive (trTIL) CD8 T cells (FIG. 24F). Differential gene expression analysis performed in the stable lesion between trTILs and bystander T cells (population covariate) showed upregulation of proinflammatory chemokines (CCL5, IL32), cytotoxic granule proteins (NKG7, GZMA), HLA class 2 molecules, and interferon-response genes, alongside with inhibitory genes (TMSB4X, LSP1 and checkpoint genes) in trTILs, consistent with local activation likely in response to persisting tumor antigens (FIG. 24G).

[0217] TCR clonotype distribution between the 2 TLs was analyzed, considering only those clonotypes with a frequency of less than 0.1% in the blood, as determined by scTCRseq. Interestingly, the 2 TLs had a different representation of top trTIL clonotypes (FIG. 24H). Backtracking these clones in the clonoSEQ dataset (FIG. 23) identified 4 patterns by unsupervised clustering (FIG. 24H): (I) clones enriched in the PB, likely representing clones that escaped the filter due to incomplete coverage of the scTCRseq dataset from blood. Of note, several of the most frequent clones in the progressing TL2 fell into this cluster, suggesting that TL2 contained mostly bystander T cells. On the contrary, clones highly represented in the stable TL1 distributed over the other clusters: (II) clonotypes exclusive to the tumor at second surgery; (III) clones temporarily emerging in the PB on day +30 following the window of conditioning-induced lymphodepletion, with some of them highly represented in the tumor at second surgery; (IV) clones highly represented in the tumor at diagnosis, most of which remained detectable in the tumor at second surgery. These data suggested that many of the top TCR clonotypes of the stable TL1 were prevalently found inside the tumor at second surgery suggesting that they might have expanded after Temferon™ treatment upon stimulation from the re-programmed TME.

[0218] At last, differentially expressed genes were extracted between the predicted trTILs from TL1 vs TL2, using the TCR clonotype as a covariate. This allowed measuring the impact of the TME on the T cell state in TL1 vs. TL2 controlling for T cell clones (e.g., cells sharing the same TCR and thus recognizing the same target antigens) provided they were present in both lesions (FIG. 241). The trTILs present in the stable lesion showed an upregulation of heat shock proteins, interferon response genes and regulators of effector cytokines and T cell activation (IFI27, CREM, RGS2), but also of genes regulating RNA metabolism and translation (SRSF3, RPS29, PCBP1), as well as genes associated withproliferation and TIL dysfunction (MTs, HIST1H4C). On the contrary, genes associated with less differentiated T cell states and absence of mutation-associated neoantigen reactivity (e.g. TXNIP, IL7R) were enriched in the putative trTILs from the progressing lesion TL2.

[0219] Taken together, these data provided an association between the Temferon™- reprogrammed myeloid TME and the accumulation of effector CD8 T cells with phenotypic features of robust activation likely against tumor associated antigens, in the context of a stringent comparison between clinically and radiologically stable vs. progressing tumor lesion, performed at the same time from the same patient. Although such analysis was only performed in one patient, these data may provide a granular representation at single-cell level of a protective innate and adaptive anti-tumor immune response induced by Temferon™5.2.1.11 Overall Survival of Patients at Second Surgery

[0220] Kaplan Meier overall survival (OS) curves were generated based on expression data collected from patients who underwent a second tumor resection surgery following Temferon™ administration. OS was calculated based on the percentage of CD8 expression in the tumor at the time of the second surgery; patients were divided into two groups: an increased CD8 expression group or a reduced CD8 expression group. The analysis found that patients with increased CD8 expression within the tumor had longer overall survival compared to patients with reduced CD8 expression within the tumor (FIG. 27).6. EQUIVALENTS AND INCORPORATION BY REFERENCE

[0221] While the invention has been particularly shown and described with reference to a preferred embodiment and various alternate embodiments, it will be understood by persons skilled in the relevant art that various changes in form and details can be made therein without departing from the spirit and scope of the invention.

[0222] All references, issued patents and patent applications cited within the body of the instant specification are hereby incorporated by reference in their entirety, for all purposes.

Claims

WHAT IS CLAIMED IS:

1. A method of treating a solid cancer, the method comprising: administering a therapeutically effective number of autologous CD34+hematopoietic stem and progenitor cells (HSPCs) to a patient with a solid cancer, wherein the HSPCs have been engineered to express interferon-a (IFN-a) from an exogenous IFN-a coding sequence that is operably linked to both positive and negative expression control elements, wherein the control positive and negative control elements together restrict expression of the exogenous IFN-a coding sequence to differentiated myeloid progeny cells within a tumor microenvironment (TME), wherein the therapeutically effective number of engineered autologous CD34+HSPCs is at least 1.5 E6 cells per kilo patient body weight (cells / kg), and administering a therapeutically effective amount of a checkpoint inhibitor.

2. The method of claim 1, wherein the positive expression control element comprises a Tie2 gene enhancer and promoter operably linked to the exogenous IFN-a coding sequence.

3. The method of claim 1 or claim 2, wherein the negative expression control element comprises at least one miRNA target sequence (mirT) engineered into the expressed IFN-a transcript, wherein each of the at least one mirT is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells.

4. The method of claim 3, wherein the negative expression control comprises a plurality of mirTs engineered into the expressed IFN-a transcript, wherein each of the plurality of mirTs is a target for a miRNA expressed in HSPCs but not in differentiated myeloid progeny cells.

5. The method of claim 3 or claim 4, wherein at least one mirT is a miR-126 target (miR-126T).

6. The method of any one of claims 3-5, wherein at least one mirT is a miR-130a target (miR-130aT).

7. The method of claim 5, wherein the negative control element comprises at least two mir-126T engineered into the IFN-a transcript.

8. The method of claim 7, wherein the negative control element consists of two mir-126T in the IFN-a transcript.

9. The method of any one of claims 1-8, wherein the HSPCs have been engineered via integration of a lentivirus expression vector.

10. The method of claim 9, wherein each of the mirT targets is encoded by a sequence in the integrated lentivirus expression vector that is transcribed into the IFN-a mRNA.

11. The method of claim 9 or claim 10, wherein the lentivirus is a replicationdefective 3rd-generation pseudotyped vector.

12. The method of claim 11, wherein the lentivirus is TIA126-LV.

13. The method of any one of claims 1-12, wherein the therapeutically effective number of engineered autologous CD34+HSPCs is at least 1.5 E6 cells / kg.

14. The method of any one of claims 1-13, wherein the therapeutically effective number of engineered autologous CD34+HSPCs is at least 2.5 E6 cells / kg.

15. The method of any one of claims claim 1-14, wherein the therapeutically effective number of engineered autologous CD34+HSPCs is at least 3.0 E6 cells / kg.

16. The method of any one of claims 1-15, wherein the therapeutically effective number of engineered autologous CD34+HSPCs is at least 3.5 E6 cells / kg.

17. The method of any one of claims 1-16, wherein the therapeutically effective number of engineered autologous CD34+HSPCs is at least 4.0 E6 cells / kg.

18. The method of claim 17, wherein the therapeutically effective number of engineered autologous CD34+HSPCs is 4E6 cells / kg.

19. The method of any one of claims 1-18, further comprising administering untransduced CD34+supporter cells in combination with the engineered CD34+HSPCs.

20. The method of claim 19, wherein 3E6 supporter cells / kg patient body weight are administered.

21. The method of claim 19, wherein 2E6 supporter cells / kg are administered.

22. The method of any one of claims 1-21, further comprising the step, preceding administration of engineered autologous CD34+cells, of: administering at least one sub-myeloablative conditioning regimen.

23. The method of claim 22, wherein the at least one sub-myeloablative conditioning regimen is selected from BCNU and thiotepa, busulfan and thiotepa, and busulfan without thiotepa.

24. The method of claim 23, wherein the at least one sub-myeloablative conditioning regimen is busulfan without thiotepa.

25. The method of any one of claims 1-24, wherein the solid cancer is glioblastoma multiforme (GBM).

26. The method of claim 25, wherein the solid cancer is GBM with unmethylated MGMT promoter (uMGMT).

27. The method of claim 25 or claim 26, further comprising the step, prior to administering the engineered autologous CD34+HSPCs, of surgically resecting the tumor mass.

28. The method of claim 27, further comprising the step, before and / or after surgical resection and before administering the engineered autologous CD34+HSPCs, of administering radiotherapy to the patient.

29. The method of any one of claims 1-28, wherein the checkpoint inhibitor is an inhibitor of PD-1, PD-L1, CTLA-4, LAG-3, ICOS, BTLA, TIM-3, TIGIT, or NKG2A.

30. The method of claim 29, wherein the checkpoint inhibitor is an antibody or antigen-binding fragment that binds specifically to PD-1, PD-L1, CTLA-4, LAG-3, ICOS, BTLA, TIM-3, TIGIT, or NKG2A.

31. The method of claim 30, wherein the checkpoint inhibitor specifically binds toPD-1.

32. The method of claim 31, wherein the PD-1 inhibitor is an anti -PD-1 antibody selected from the group consisting of: pembrolizumab, nivolumab, and cemiplimab.

33. The method of claim 32, wherein the anti-PD-1 antibody is pembrolizumab.

34. The method of claim 32, wherein the anti-PD-1 antibody is nivolumab.

35. The method of claim 32, wherein the anti-PD-1 antibody is cemiplimab.

36. The method of claim 30, wherein the checkpoint inhibitor specifically binds toPD-L1.

37. The method of claim 36, wherein the PD-L1 inhibitor is an anti-PD-Ll antibody selected from the group consisting of: atezolizumab, avelumab, and durvalumab.

38. The method of claim 30, wherein the checkpoint inhibitor specifically binds to CTLA-4.

39. The method of claim 38, wherein the CTLA-4 inhibitor is an anti-CTLA-4 antibody selected from the group consisting of: ipilimumab and tremelimumab.

40. The method of claim 30, wherein the checkpoint inhibitor specifically binds to LAG-3.

41. The method of claim 40, wherein the LAG-3 inhibitor is the anti -LAG-3 antibody is relatlimab.

42. The method of claim 30, wherein the checkpoint inhibitor specifically binds to ICOS.

43. The method of claim 42, wherein the anti-ICOS antibody is MEDI-570.

44. The method of claim 30, wherein the checkpoint inhibitor specifically binds to BTLA.

45. The method of claim 30, wherein the checkpoint inhibitor specifically binds to TIM-3.

46. The method of claim 30, wherein the checkpoint inhibitor specifically binds to TIGIT.

47. The method of claim 46, wherein the anti-TIGIT antibody is selected from vibostolimab, etigilimab, and tiragolumab.

48. The method of claim 30, wherein the checkpoint inhibitor specifically binds to NKG2A.

49. The method of any one of claims 1-48, wherein the checkpoint inhibitor is administered prior to, concurrently with, and / or subsequent to administration of the autologous CD34+HSPCs.

50. The method of claim 49, wherein the checkpoint inhibitor is first administered after confirmation of HSPC engraftment.

Images