Precise combination therapy of radiotherapy and immunotherapy guided by joint PD-L1 imaging and T-cell imaging.

BACKGROUND: PD-1/PD-L1 blockade therapy is often used in combination with radiotherapy (RT) to enhance therapeutic efficacy in clinic. However, the effectiveness of the combination therapy is still unsatisfactory due to the blindness in the administration schedule. More seriously, multiple dosing-induced deaths of PD-1/PD-L1 blockade often occur. Considering the urgency to enhance the efficacy and safety of the combination therapy, we pioneered a strategy that used SPECT/CT imaging of tumor PD-L1 expression and tumor-infiltrating T cells to guide and predict the combination therapy toward precision cancer therapy. METHODS: Single-photon emission computed tomography (SPECT)/CT imaging of a PD-L1-targeting probe (99mTc-MY1523) was employed to monitor the dynamic changes of tumor PD-L1 expression following RT, guiding the timing of αPD-1 therapy. SPECT/CT imaging of a T-cell-targeting probe (99mTc-sum IL-2) was employed to predict the prognosis of combination therapy. The findings were validated in both immune hot and cold tumor models, using high-dose single RT or fractionated RT. RESULTS: As indicated by the dynamic imaging of tumor PD-L1 expression, the combination therapy achieved optimal antitumor efficacy, only when αPD-1 antibody was administrated during the window period of PD-L1 upregulation following RT. However, although PD-L1 imaging could indicate when the combination therapy was more effective, it could not accurately predict the prognosis of the combination therapy. The T-cell imaging showed that the changed number of tumor-infiltrating T cells, rather than the absolute number at a certain time, could predict the prognosis in the early stage. The single-cell RNA sequencing analysis showed that the αPD-1 therapy after RT during the time window could prevent or reverse the exhausted CD8+ T cell state and switched the tumor immune microenvironment to an antitumor state, only in 'responsive tumors' (imaging showed an increase in tumor-infiltrating T cells) rather than 'non-responsive tumors'. CONCLUSIONS: PD-L1 imaging was suitable for guiding the timing of αPD-1 administration, while T-cell imaging was more suitable for predicting prognosis. The two imaging modalities jointly provide essential guidance to enhance the efficacy and safety of the combination therapy, facilitating individualized precision therapy. This imaging strategy can also be extrapolated to clinical practice for immunotherapy combined with other therapies to boost efficacy.