So first, they are target and drug-related mechanisms. These include variability in PSMA expression, as well as changes in ligand binding, trafficking, and overall pharmacokinetics. Second, we have evolutionary and clonal mechanisms. So under the selective pressure of PSMA-RLT, you can select a population that are PSMA-negative, which ultimately will expand and lead to disease relapse. Third, you have microenvironmental and systemic factors, which play a critical role. So tumor stroma, abnormal vasculature, hypoxia, physical barriers can limit the radiation penetrance into the tissue. And finally, DNA damage and repair mechanisms are modulated by the tumor cells in response to PSMA-RLT. So, so far, the clinical experience and emerging evidence suggests that the dominant reason for poor response to PSMA-RLT is the presence of PSMA-negative disease, either preexisting or therapy-induced.
So today I will attempt to discuss the first three mechanisms as DNA damage has been already largely covered in that session. You've probably all have seen this study, but first, if we focus on lesions and their uptake, response to PSMA-RLT does not depend just on uptake, but on how uniform that uptake is across the disease. High homogeneous PSMA uptake is associated with the longest survival, while heterogeneous and low PSMA uptake predict poor response and poor outcome. So while SUV stratification correlates with survival, high SUV combined with heterogeneous PSMA expression drives mixed response, early relapse, and overall survival. So this highlights that lesion-level heterogeneity is a key mechanism of resistance to PSMA-RLT. We actually modeled that in our group in a preclinical study, and we showed that variable PSMA uptake was observed across the heterogeneous tumor, which directly translated in different DNA damage induction and ultimately therapeutic efficacy reflective of the inherent heterogeneity. Clinically, what does this mean? It just means that PSMA PET positivity alone is not enough.
It's important to consider the degree of PSMA expression, the spatial distribution, and how consistently it is expressed across all tumor types. So taken together, like in those two slides, we can see that heterogeneity limits uniform dose delivery and creates a biological foundation for treatment resistance and eventual relapse. Beyond heterogeneity in PSMA, the underlying biology of the non-PSMA tumor plays an equally important role in potential relapse. As such, in a retrospective imaging analysis of 102 patients, combined analysis of PSMA PET and FDG-PET were used for staging and for follow-up after two cycles of PSMA, and it shows that these tumors have undergone a glucometabolic change. So low changes in whole-body lesion glycolysis normalized to whole-body PSMA expression was associated with significantly longer survival, hinting that PSMA PET alone may underestimate disease burden while FDG uptake unmasks emerging resistant clones. Overall, the dynamic metabolic shift in the tumor, not just baseline PSMA expression, appears to be a critical determinant of outcome.
As you probably already know, prostate cancer is rarely a uniform disease and lineage plasticity may represent a crucial evolutionary mechanism driving resistance to PSMA-based radioligand therapy. So in these two cohorts of patients from the Beltran research group or the autopsy trial from Michael Haffner's group, PSMA clonal heterogeneity is present across all stages of prostate cancer and varies within lesions. Under therapeutic pressure such as PSMA-RPT, prostate cancer can undergo lineage switching most notably towards a neuroendocrine-like phenotype. So these NEPC are characterized by downregulation of PSMA and upregulation of glycolytic and metabolic pathways, including increased expression of glucokinase. Clinically, what does this mean? This may manifest as PSMA-FDG discordance or increased FDG avidity, signaling an emerging treatment-resistant phenotype. This metabolic reprogramming may reflect a broader shift of PSMA dependence towards glucose-driven survival mechanisms. So essentially, you have a switch between PSMA and glucose metabolism. Beyond tumor-intrinsic mechanisms, an important contributor to resistance can also be the tumor microenvironment, so the TME shapes response to PSMA-RPT by limiting drug delivery, altering immunity, and promoting radioresistance.
Then stroma immunosuppressive cells, abnormal vessels, and interstitial pressure and hypoxia limit PSMA penetration and effectiveness. You have also the possibility to see downregulation of DNA sensing mechanisms which blunt the immune activation following radioligand therapy. So in normal condition following DNA damage by radiation, the activation of the STING pathway will lead to the expression of interferon-stimulating genes and the production of type I interferon, which will lead to the activation of effector immune cells. In prostate cancer, STING and interferon signaling are downregulated in patients which lead to a blunted immune response. We wanted to assess that in mice to understand whether what's the downregulation of the STING in patients lead in terms of response to PSMA-RLT. So we modeled both extremes of STING expression using STING-competent or STING-deficient tumors while maintaining STING signaling. In STING-positive tumors, as you can see on the upper panel, PSMA-RPT led to prolonged therapeutic benefits, which extended to 49 days. In contrast, STING-negative tumors showed reduced response resulting in shorter survival. So this finding highlights that tumor-intrinsic STING status is a key determinant of RPT response, helping explain variability in patient response despite similar radiation exposure.
So I'll close my talk by summarizing the key concepts of radioresistance and relapse occur after PSMA-RLT and what this means for future treatment strategy. So as I've explained, one of the most dominant drivers of poor response is heterogeneity in PSMA expression, and it's important to shift tumors towards a more homogeneous membrane-localized PSMA expression profile to improve radiation delivery and efficacy. As tumors evolve as well, non-PSMA-expressing clones often expand. This suggests that PSMA-dependent therapy alone will be insufficient and that antigen-agnostic or non-PSMA strategies must be considered earlier in the disease course. The TME has played as well a critical role in limiting response. Strategies that remodel the TME or restore innate immune sensing may enhance PSMA-RPT efficacy. This observation supports combining PSMA-RPT with immune-based approaches to sustain and deepen the immune response. In summary, the most likely reason for poor response to lutetium-PSMA, relapse, and poor survival is the coexistence of PSMA-negative disease. And I hope that throughout my presentation, despite all the stress, that I've hammered that prostate cancer is rarely a uniform disease, it is important to integrate PSMA, FDG, and other imaging modalities to enable a more complete view of the disease biology beyond a single snapshot.
Combining multiplex imaging with systems analysis of the tumor can reveal metabolic diversity, PSMA expression patterns, and the distinct tumor subtypes. This is the concept of personomics. It will be powerful and it will tailor treatment, not just to the average biology of the tumor, but to the unique multidimensional phenotype of each patient's disease. In practical terms, it means better patient selection, earlier identification of non-responders, and rational design of combinations. Ultimately, phenotyping brings us closer to precision PSMA-RPT, where therapy is adapted to the individual patient. Sorry. In summary, to fully leverage the potential of PSMA therapy, we must go beyond PSMA. Thank you.