Diabetes Mellitus Is a Strong Independent Negative Prognostic Factor in Patients with Brain Metastases Treated with Radiotherapy

SIMPLE SUMMARY: Cerebrovascular disorders are common among cancer patients. They might influence tumor growth, treatment sensitivity and, ultimately, the prognoses of patients with brain metastases (BM). In a retrospective exploratory study, we examined if the presence of arterial hypertension, smoking, diabetes mellitus (DM), hypercholesterolemia or peripheral arterial occlusive disease has a prognostic impact in patients with BM. In uni- and multivariate analysis, the presence of DM was associated with a worse prognosis across several tumor types, while for the other cerebrovascular risk factors, significant differences in survival were not found. From molecular data, it can be hypothesized that RAGE activation plays an important role in the interaction between DM and BM. In future studies, it remains to be determined to what extent serum glucose levels and antidiabetic treatments may influence survival and if optimized antidiabetic treatment or RAGE targeted treatments are able to improve prognoses of patients with BM. ABSTRACT: Background: Brain metastases (BM) cause relevant morbidity and mortality in cancer patients. The presence of cerebrovascular diseases can alter the tumor microenvironment, cellular proliferation and treatment resistance. However, it is largely unknown if the presence of distinct cerebrovascular risk factors may alter the prognosis of patients with BM. Methods: Patients admitted for the radiotherapy of BM at a large tertiary cancer center were included. Patient and survival data, including cerebrovascular risk factors (diabetes mellitus (DM), smoking, arterial hypertension, peripheral arterial occlusive disease, hypercholesterolemia and smoking) were recorded. Results: 203 patients were included. Patients with DM (n = 39) had significantly shorter overall survival (OS) (HR 1.75 (1.20–2.56), p = 0.003, log-rank). Other vascular comorbidities were not associated with differences in OS. DM remained prognostically significant in the multivariate Cox regression including established prognostic factors (HR 1.92 (1.20–3.06), p = 0.006). Furthermore, subgroup analyses revealed a prognostic role of DM in patients with non-small cell lung cancer, both in univariate (HR 1.68 (0.97–2.93), p = 0.066) and multivariate analysis (HR 2.73 (1.33–5.63), p = 0.006), and a trend in melanoma patients. Conclusion: DM is associated with reduced survival in patients with BM. Further research is necessary to better understand the molecular mechanisms and therapeutic implications of this important interaction.