Neoantigen Cancer Vaccine Patents

Neoantigen cancer vaccine patents cover neoantigen-identification and prediction innovations; personalized-vaccine construct and manufacturing innovations; shared/off-the-shelf-antigen innovations; and combination and delivery innovations — with IP held by mRNA leaders, personalized-vaccine pioneers, and neoantigen-prediction specialists. MAJOR NEOANTIGEN-VACCINE PATENT HOLDERS: MODERNA + MERCK: mRNA-4157/V940, an individualized neoantigen therapy (an mRNA encoding up to ~34 patient-specific neoepitopes) combined with Keytruda (pembrolizumab), with positive melanoma Phase 2 and ongoing Phase 3 — a leading personalized-vaccine program. BIONTECH: autogene cevumeran (individualized neoantigen mRNA, with Genentech/Roche, notably in pancreatic cancer), the iNeST individualized platform, and FixVac (off-the-shelf shared tumor antigens). GRITSTONE BIO: the EDGE machine-learning neoantigen-prediction model and self-amplifying-mRNA and viral-vector (ChAd) delivery. OTHERS: Genentech/Roche (BioNTech partner), Transgene/NEC (TG4050 individualized), Nouscom (off-the-shelf neoantigen for MSI tumors), Geneos, Genocea heritage, and academic founders (the Dana-Farber/Broad and Washington University neoantigen-prediction work). Neoantigen prediction, personalized construct/manufacturing, and checkpoint combination are the core neoantigen-vaccine patent domains.

Tumor-sequencing and variant-calling innovations; HLA-typing and MHC-binding prediction innovations; epitope-ranking and selection innovations; and immunogenicity-prediction innovations represent core neoantigen-vaccine patent domains — though prediction-algorithm claims face §101 scrutiny and are strongest tied to the wet-lab/manufacturing pipeline. SEQUENCING / VARIANT PATENTS: tumor-versus-normal whole-exome and RNA sequencing to find somatic mutations, variant calling and expression filtering, and identifying tumor-specific mutated peptides (neoantigens) plus tumor-specific splice/fusion/indel neoantigens. HLA / BINDING-PREDICTION PATENTS: patient HLA typing, MHC class I/II binding prediction, and antigen-processing/presentation prediction (proteasomal cleavage, TAP transport) — the core computational step. RANKING / SELECTION PATENTS: machine-learning models that rank candidate neoepitopes by predicted immunogenicity (Gritstone EDGE, Moderna/BioNTech proprietary models trained on mass-spec immunopeptidomics and T-cell-response data), and selecting the best ~20–34 epitopes for the vaccine. IMMUNOGENICITY PATENTS: mass-spectrometry immunopeptidomics (directly observing presented peptides) and T-cell-reactivity validation. These prediction methods are most defensible claimed together with the specific sequencing, mass-spec, and manufacturing pipeline (a concrete technical process) rather than as abstract algorithms — the model itself is often kept as a trade secret. Accurate neoantigen prediction is the highest-value technical differentiator (better predictions → more immunogenic vaccines).

Personalized-manufacturing and logistics innovations; vaccine-construct and delivery innovations; shared/off-the-shelf-antigen innovations; and combination innovations represent additional neoantigen-vaccine patent domains — and rapid per-patient manufacturing is the operational heart of personalized cancer vaccines. MANUFACTURING PATENTS: the rapid, individualized 'N-of-1' manufacturing process (sequence tumor → predict → design → synthesize → release a bespoke vaccine in weeks), automated/parallelized GMP production, quality control and release for a one-patient batch, and supply-chain/logistics — this manufacturing IP is central and defensible because personalization is an operational, not just scientific, problem. CONSTRUCT / DELIVERY PATENTS: the vaccine modality encoding the neoepitopes — mRNA-LNP (concatenated neoepitope-encoding mRNA in lipid nanoparticles, sharing LNP IP with mRNA vaccines), synthetic long peptides, viral vectors (chimp adenovirus ChAd + self-amplifying-mRNA prime/boost — Gritstone), and DNA; plus epitope-linker/cassette design and self-amplifying RNA. SHARED-ANTIGEN PATENTS: off-the-shelf vaccines targeting recurrent shared neoantigens (KRAS G12D/G12C, TP53 hotspots, MSI frameshift neoantigens) — composition-of-matter on defined antigen sets, avoiding per-patient manufacturing. COMBINATION PATENTS: combination with checkpoint inhibitors (the clinical standard — vaccine + anti-PD-1) and other immunotherapies, and dosing regimens. Personalized manufacturing and the mRNA-LNP neoepitope construct are the highest-value neoantigen-vaccine IP.

Neoantigen vaccine startup IP strategy must navigate Moderna/Merck and BioNTech personalized-platform and manufacturing patents, mRNA and LNP IP (shared with mRNA vaccines — Moderna/BioNTech/Acuitas/Arbutus), Gritstone prediction/delivery patents, a §101 constraint on pure prediction algorithms, and a landscape where prediction accuracy and manufacturing logistics — not just the construct — decide success; understand that the durable IP is in the manufacturing process, the vaccine construct/delivery, and shared-antigen compositions, that neoantigen-prediction models are §101-sensitive and often best kept as trade secrets (the training data and model are the moat), and that LNP/mRNA freedom-to-operate is a gating issue; identify whitespace in shared/off-the-shelf neoantigens, improved prediction (tied to wet-lab), non-LNP or extrahepatic delivery, and rapid manufacturing. NEOANTIGEN-VACCINE STARTUP IP STRATEGY: MANUFACTURING, CONSTRUCT, AND SHARED ANTIGENS ARE THE IP — PREDICTION IS OFTEN TRADE-SECRET: patent the rapid N-of-1 manufacturing process, the neoepitope construct/delivery, and shared-antigen compositions; keep the neoantigen-prediction model and training data as trade secrets (and claim prediction only when tied to the concrete pipeline, given §101); SHARED/OFF-THE-SHELF NEOANTIGENS ARE HIGHEST-VALUE WHITESPACE: defined recurrent neoantigens (KRAS/TP53 hotspots, MSI frameshifts) give composition-of-matter without per-patient manufacturing — a major commercial and patent opportunity; PREDICTION ACCURACY IS THE SCIENTIFIC MOAT (KEEP IT SECRET): the model trained on immunopeptidomics + T-cell data is the differentiator — protect via trade secret and data scale, not disclosure; LNP/mRNA FTO IS GATING: personalized mRNA vaccines rely on LNP/mRNA IP (Moderna/BioNTech/Acuitas/Arbutus) — secure freedom-to-operate or use alternative modalities (peptide, viral vector); MANUFACTURING LOGISTICS IS A REAL MOAT: weeks-to-patient bespoke GMP production is an operational barrier worth patenting; WHEN TO PATENT: NOVEL CONSTRUCT/PROCESS WITH MEASURED IMMUNOGENICITY: file once a candidate shows measured results (neoantigen prediction precision/recall + T-cell response rate + manufacturing turnaround + clinical response with checkpoint) — measured prediction accuracy, T-cell immunogenicity, manufacturing speed, and clinical benefit are the critical neoantigen-vaccine IP metrics; KEY FTO CHECKLIST: Moderna/Merck mRNA-4157 individualized neoepitope mRNA-LNP + Keytruda; BioNTech iNeST autogene cevumeran, FixVac shared-antigen; Gritstone EDGE prediction, ChAd + self-amplifying-mRNA prime/boost; tumor/normal exome+RNA-seq variant calling; HLA typing MHC-binding prediction; mass-spec immunopeptidomics; N-of-1 GMP manufacturing/release; KRAS/TP53/MSI shared neoantigens; LNP/mRNA FTO (Acuitas/Arbutus/Genevant); checkpoint-inhibitor combination; prediction §101-tied-to-pipeline/trade-secret.