Optimizing Engraftment of Gastric Cancer Patient-derived Xenograft Models and Monitoring with MRI

Gastric cancer (GC) Patient-derived xenograft PDX models have low engraftment rates (12–15%). This study aims to examine the critical factors and evaluation approaches in GC PDX models. Tumor tissues from 33 GC patients were transplanted into NOD/SCID mice. Clinical and pathological factors, tumor sampling methods, implant site, tissue size, tissue pathological type, tissue differentiation degree and out-of-body time were analyzed. Successful engraftment (38.4%, 28/73 models) correlated with higher albumin ( p  = 0.007), factor VII ( p  = 0.009), lymphocyte count ( p  = 0.023), and lower neutrophil-to-lymphocyte ratio ( p  = 0.022). Optimal conditions included surgical samples ( p  = 0.007), poorly differentiated tumors ( p  = 0.028), 2×2×2 mm³ tissue size ( p  = 0.033), renal capsule transplantation ( p  = 0.001), and ≤ 2-hour processing ( p  = 0.003). MRI tracked tumor dynamics, showing initial shrinkage before growth in successful models. PDX tumors closely mirrored primary tumors histologically and genetically. Key improvements include using 2×2×2 mm³ surgical samples transplanted into renal capsules within 2 hours. Patient albumin, absolute lymphocyte count, factor VII, and neutrophil-to-lymphocyte ratio levels predict success, and MRI enables early monitoring. These refinements significantly enhance GC PDX engraftment.