Background Most glioblastomas recur near prior radiation treatment sites. Future clinical success will require achieving and optimizing an “abscopal effect,” whereby unirradiated neoplastic cells outside treatment sites are recognized and attacked by the immune system. Radiation combined with anti–programmed cell death ligand 1 (PD-L1) demonstrated modest efficacy in phase II human glioblastoma clinical trials, but the mechanism and relevance of the abscopal effect during this response remain unknown. Methods We modified an immune-competent, genetically driven mouse glioma model (forced platelet derived growth factor [PDGF] expression + phosphatase and tensin homolog loss) where a portion of the tumor burden is irradiated (PDGF) and another unirradiated luciferase-expressing tumor (PDGF + luciferase) is used as a readout of the abscopal effect following systemic anti–PD-L1 immunotherapy. We assessed relevance of tumor neoepitope during the abscopal response by inducing expression of epidermal growth factor receptor variant III (EGFRvIII) (PDGF + EGFRvIII). Statistical tests were two-sided. Results Following radiation of one lesion, anti–PD-L1 immunotherapy enhanced the abscopal response to the unirradiated lesion. In PDGF-driven gliomas without tumor neoepitope (PDGF + luciferase, n = 8), the abscopal response occurred via anti–PD-L1 driven, extracellular signal-regulated kinase–mediated, bone marrow–derived macrophage phagocytosis of adjacent unirradiated tumor cells, with modest survival implications (median survival 41 days vs radiation alone 37.5 days, P = 0.03). In PDGF-driven gliomas with tumor neoepitope (PDGF + EGFRvIII, n = 8), anti–PD-L1 enhanced abscopal response was associated with macrophage and T-cell infiltration and increased survival benefit (median survival 36 days vs radiation alone 28 days, P = 0.001). Conclusion Our results indicate that anti–PD-L1 immunotherapy enhances a radiation- induced abscopal response via canonical T-cell activation and direct macrophage activation in glioblastoma.

PSII undergoes photodamage, which results in photoinhibition-the light-induced loss of photosynthetic activity. The main target of damage in PSII is the reaction center protein D1, which is buried in the massive 1.4 MDa PSII holocomplex. Plants have evolved a PSII repair cycle that degrades the damaged D1 subunit and replaces it with a newly synthesized copy. PSII core proteins, including D1, are phosphorylated in high light. This phosphorylation is important for the mobilization of photoinhibited PSII from stacked grana thylakoids to the repair machinery in distant unstacked stroma lamellae. It has been recognized that the degradation of the damaged D1 is more efficient after its dephosphorylation by a protein phosphatase. Recently a protein phosphatase 2C (PP2C)-type PSII core phosphatase (PBCP) has been discovered, which is involved in the dephosphorylation of PSII core proteins. Its role in PSII repair, however, is unknown. Using a range of spectroscopic and biochemical techniques, we report that the inactivation of the PBCP gene affects the growth characteristic of plants, with a decreased biomass and altered PSII functionality. PBCP mutants show increased phosphorylation of core subunits in dark and photoinhibitory conditions and a diminished degradation of the D1 subunit. Our results on D1 turnover in PBCP mutants suggest that dephosphorylation of PSII subunits is required for efficient D1 degradation.

Glioblastomas are highly malignant brain tumors. Knowledge of growth rates and growth patterns is useful for understanding tumor biology and planning treatment logistics. Based on untreated human glioblastoma data collected in Trondheim, Norway, we first fit the average growth to a Gompertz curve, then find a best fitted white noise term for the growth rate variance. Combining these two fits, we obtain a new type of Gompertz diffusion dynamics, which is a stochastic differential equation (SDE). Newly collected untreated human glioblastoma data in Seattle, US, re-verify our model. Instead of growth curves predicted by deterministic models, our SDE model predicts a band with a center curve as the tumor size average and its width as the tumor size variance over time. Given the glioblastoma size in a patient, our model can predict the patient survival time with a prescribed probability. The survival time is approximately a normal random variable with simple formulas for its mean and variance in terms of tumor sizes. Our model can be applied to studies of tumor treatments. As a demonstration, we numerically investigate different protocols of surgical resection using our model and provide possible theoretical strategies.

The aging of the western population and the increased use of oral anticoagulation (OAC) and antiplatelet drugs (APD) will result in a clinical dilemma on how to balance the recurrence risk of chronic subdural hematoma (cSDH) with the risk of withholding blood thinners. Objective: To identify features that predicts recurrence, thromboembolism (TEE), hospital stay and mortality. To identify the optimal window for resuming APD or OAC. Methods: We performed a retrospective multivariate analysis of a prospectively collected database. We then build machine learning models for outcomes prediction. Results: We identified 596 patients. The rate of recurrence was 22.17%, that of thromboembolism was 0.9% and that of mortality was 14.78%. Smoking, platelet dysfunction, CKD, and alcohol use were independent predictors of higher recurrence, while resolution of the SDH was protective. OAC use had higher odds of developing TEEs. CKD, developing a new neurological deficit or a TEEs were independent predictors of higher mortality. We find the optimal time of resuming OAC to be after 2 days but before 21 days as these patients had the lowest recurrence of bleeding associated with a low risk of stroke. The ML model achieved an accuracy of 93, precision of 0.84 and recall of 0.80 for recurrence prediction. ML models for hospital stay performed poorly (R 2 = 0.33). ML model for stroke was overfitted given the low number of events. Conclusion: ML modeling is feasible. However, large well-designed prospective multicenter studies are needed for accurate ML so that clinicians can balance the risks of recurrence with the risk of TEEs, especially for high-risk anticoagulated patients.