In the laboratory, we can cure cancer and prevent infections with remarkable efficiency. Yet, when these same therapies move from mice to humans, they often falter. This "translational gap" is not merely a matter of scale; it is a fundamental divergence in biological complexity. Across oncology and vaccinology, researchers are identifying why standard preclinical models fail to predict human outcomes and are developing strategies to bridge this divide.
The Divergent Immune Landscape
The most significant hurdle in translating immunotherapies is the fundamental difference between the murine and human immune systems. While mice provide a convenient "living test tube," they often lack the functional parity required to predict human responses to complex interventions like viral vectors or vaccines.
Eunkyung Park and colleagues demonstrated that while adenoviral vectors carrying IL-12 can trigger robust antitumor immunity in mice, these results are difficult to project onto humans because the two species react differently to viral delivery systems. Similarly, Peter B. Gilbert found that a vaccine component that appeared protective in small animal models failed completely in human clinical trials. The human immune landscape is shaped by years of "immunological memory" from prior infections—a factor Cheng Gao notes is entirely absent in standardized laboratory animals like ferrets or mice.
Furthermore, specific molecular pathways differ. Soon Park highlights that discrepancies between human and rodent complement systems—a critical arm of innate immunity—can render immune-modulating therapies that work in the lab ineffective in the clinic.
Heterogeneity Versus Standardization
In a lab, researchers often use "syngeneic" or "xenograft" models, where identical tumor cells are implanted into identical mice. This creates a clean, reproducible experiment, but it ignores the chaotic reality of human disease.
Human tumors are highly heterogeneous, meaning they contain a diverse mix of cells with different genetic mutations and physical characteristics. Suke Deng and Ying‐Tzu Yen argue that this variability is a primary reason why nanomedicines and immunotherapies that successfully "remodel" the tumor microenvironment in mice fail to do so in humans. A therapy might penetrate a uniform mouse tumor but be blocked by the "physical walls" and complex signaling of a diverse human tumor.
The Importance of Physiological Context
Where a tumor grows matters as much as what it is. For years, the standard was the "subcutaneous" model—injecting cancer cells under a mouse's skin. However, Rui Han showed that "orthotopic" models, which place tumors in their organ of origin (like the adrenal gland for neuroblastoma), produce metastatic patterns and gene signatures that far more closely resemble human disease.
Physical barriers also differ by species. In the treatment of brain cancer, Yanhong Ren points out that the human blood-brain barrier is significantly more robust than that of a mouse. A nanoparticle that easily reaches a glioma in a preclinical model may never reach the target in a patient, leading to a false sense of efficacy during early development.
Engineering Better Bridges
To solve these problems, the field is moving toward more sophisticated, "humanized" environments:
- Humanized Mice: Asen Bagashev and others are using mice engineered with human immune cells to test therapies like in vivo CAR-macrophages, though even these models have inherent biological limits.
- Nonhuman Primates (NHPs): For vaccines, Joseph R. Francica emphasizes that NHPs are essential for capturing the innate transcriptional effects that mice simply cannot replicate.
- Advanced Delivery: Leland Metheny suggests that changing the delivery route—such as intra-osseous injection directly into the bone marrow—can overcome some of the engraftment failures seen in traditional intravenous models.
By acknowledging that a mouse is not a small human, researchers like Diya Zhang are shifting the focus from high-throughput testing to high-fidelity modeling, ensuring that the "cures" found in the lab have a fighting chance in the clinic.