Clinical Research Directory

Clinical Study of the Therapeutic Effectiveness of In-silico-Designed, Machine Learning Inspired, and Quantum-molecularly Coupled Personalized Neoantigenic Vaccines Microlyvaq™ in Patients With Advanced Non-small Cell Lung Cancer

Microlyvaq™ is a first-line, non-randomized, two-arm clinical trial in advanced non-small cell lung cancer (NSCLC). In both arms, patients receive a personalized multi-epitope vaccine (Microlyvaq™) on top of standard-of-care chemo-immunotherapy, with treatment tailored by histology: Arm 1 - Squamous NSCLC: Microlyvaq™ + carboplatin AUC 5 + paclitaxel 175 mg/m² + pembrolizumab Arm 2 - Non-squamous NSCLC: Microlyvaq™ + carboplatin AUC 5 + pemetrexed 500 mg/m² + pembrolizumab Because this is a non-randomized study, patients are assigned to arms based on tumor histology (squamous vs non-squamous), not by random allocation. The core problem it addresses is that even with pembrolizumab plus histology-appropriate chemotherapy, many patients either never respond or respond briefly and then progress. Tumors evade by exhausting T cells, excluding them from the tumor bed, evolving antigen loss, and maintaining suppressive myeloid and stromal niches. Microlyvaq™ is designed to overcome these resistance modes by actively installing new, durable, polyfunctional anti-tumor immunity rather than relying only on pre-existing T cells. Here's how it works. Each patient's tumor is sequenced (whole exome and RNA-seq) to identify both well-known lung cancer-associated antigens (e.g. NY-ESO-1, SOX2, p53, MAGE-A4, BRAF, BMI1, FXR1, HuD, HuC, CAGE) and private neoantigens created by that tumor's specific mutations, fusions, and splice variants. From this large antigen pool, machine learning models score each candidate epitope for that specific patient. The models consider predicted HLA class I and II presentation, how efficiently the antigen will actually be processed and displayed, whether it's expressed in tumor but not healthy tissue, how essential it is to most malignant cells (to avoid easy escape), and whether it is likely to drive functional, non-exhausted T-cell responses. This is not a generic ranking; it is individualized per patient. The most promising epitopes then undergo a quantum molecular coupling evaluation. Instead of simply asking whether a peptide binds a given HLA, Microlyvaq™ modeling simulates the peptide-MHC complex as a physical system and approximates solutions to Ĥψ = Eψ to estimate whether the peptide will form a stable, low-energy, presentation-competent conformation that a realistic T-cell receptor can dock to without high energetic penalty. Epitopes that look good in simple binding screens but are predicted to be unstable, transient, or geometrically inaccessible to TCRs are excluded. The remaining epitope set is engineered to: (1) recruit potent CD8⁺ cytotoxic T cells that can kill tumor cells, and (2) recruit CD4⁺ Th1 helper T cells that produce IFN-γ, TNF-α, and IL-2 to sustain and support those killers. The vaccine is therefore intentionally multi-epitope, Th1-biased, and patient-specific. Each personalized Microlyvaq™ lot is manufactured under GMP and given as a prime-boost series in sync with pembrolizumab and the appropriate chemotherapy backbone for the patient's histologic arm (carboplatin/paclitaxel for squamous; carboplatin/pemetrexed for non-squamous). Timing is deliberate: chemotherapy induces immunogenic tumor cell death and antigen release and transiently "opens up" the tumor microenvironment, while pembrolizumab lifts PD-1-mediated brakes on emerging T cells. Microlyvaq™ is dosed into that vulnerable window to expand vaccine-encoded clones just as new antigen is exposed and suppression is partially relieved. The goal is to generate rapid tumor shrinkage, then sustained immune pressure on residual disease, plus epitope spreading - where the immune system begins to recognize additional tumor targets beyond those in the vaccine, making escape more difficult. The trial itself is structured as a seamless, adaptive, non-randomized Phase I/IIa study, with two predefined histology-based arms (squamous vs non-squamous) rather than randomized treatment allocations. The primary early endpoint is objective response rate (RECIST v1.1). Key secondary endpoints include progression-free survival, duration of response, and overall survival. In addition, the study incorporates real-time translational signals as decision points, including: 1. polyfunctional Th1 and CD8⁺ responses to vaccine epitopes by ELISpot/ICS, 2. durable expansion and persistence of vaccine-linked TCR clonotypes in blood and, when feasible, in tumor, 3. rapid decline in circulating tumor DNA as an early molecular marker of tumor clearance, 4. improved tumor infiltration by CD8⁺ and Th1 cells, and 5. remodeling of the tumor microenvironment away from suppressive myeloid states. If a given histology arm shows strong clinical responses plus these immune/molecular signals, that arm can seamlessly expand into survival-powered confirmation. If it does not, predefined futility rules allow that arm to stop, all within this non-randomized, adaptive framework.