Hypothesis

Pretreatment detection of phosphorylated LDHA (Y10) coupled with elevated tumor lactate concentrations predicts non‑response to anti‑PD1 immunotherapy in melanoma, independent of genomic alterations.

Mechanistic Rationale

LDHA phosphorylation at Y10 stabilizes the enzyme, boosting glycolytic flux and lactate excretion [1]. High extracellular lactate acidifies the microenvironment, inhibiting cytotoxic T‑cell function through HIF1α‑driven upregulation of PD‑1 and exhaustion markers [2]. Proteogenomic studies show that protein‑level alterations, such as LDHA phosphorylation, are often missed at the transcript level but directly affect drug response [3]. Thus, a proteo‑metabolic signature of LDHA activity captures the functional state that determines immune evasion.

Experimental Design

1. Cohort: Enroll 120 treatment‑naïve melanoma patients scheduled for anti‑PD1 therapy; collect core biopsy before treatment.
2. Multi‑omics profiling:
   • Phospho‑proteomics (TiO2 enrichment, LC‑MS/MS) to quantify LDHA Y10 phosphorylation.
   • Targeted metabolomics (LC‑MS) for lactate, pyruvate, and TCA intermediates.
   • Spatial transcriptomics (Visium) to map CD8‑cell IFNγ signature and exhaustion markers (PD‑1, TIM‑3, LAG‑3).
   • Whole‑exome sequencing to control for genomic covariates (e.g., PTEN loss, MAPK mutations).
3. Model building: Generate a logistic regression model using pLDHA Y10 intensity, lactate concentration, and CD8 IFNγ score as predictors of clinical response (RECIST) at 12 weeks.
4. Validation: Split data 70/30 for training/testing; assess model performance via AUC, calibration, and net reclassification improvement over a genomics‑only baseline.
5. Mechanistic assay: Ex vivo co‑culture of patient‑derived tumor slices with autologous T‑cells; treat with LDHA inhibitor (GSK2837808A) and measure lactate, pH, and T‑cell cytotoxicity.

Expected Outcomes

• Patients with high pLDHA Y10 (>75th percentile) and lactate (>5 mM) will exhibit significantly lower objective response rates (ORR) compared to those with low levels (expected ORR 20% vs 55%).
• The proteo‑metabolic model will achieve an AUC ≥0.80, outperforming the genomics‑only model (AUC ≈0.65).
• LDHA inhibition in expl cultures will reduce lactate by ≥40%, increase pH, and restore IFNγ production in CD8‑T cells, supporting causality.

Potential Pitfalls and Mitigations

• Tumor heterogeneity: Use multiple core biopsies and spatial transcriptomics to capture regional variation.
• Technical variability: Apply batch‑correction methods (ComBat) and include internal standards for phospho‑peptide quantification.
• Confounding therapies: Exclude patients receiving neoadjuvant therapy or steroids prior to biopsy.

If the model fails to improve prediction over genomics alone, the hypothesis is falsified, indicating that LDHA phosphorylation and lactate are not dominant drivers of immune checkpoint resistance in this context.