Abstract Background Accumulation of malignant plasma cells in the bone marrow causes lytic bone lesions in 80% of multiple myeloma patients. Frequently fracturing, they are challenging to treat surgically. Myeloma cells surviving treatment in the presumably protective environment of bone lesions impede their healing by continued impact on bone turnover and can explain regular progression of patients without detectable minimal residual disease (MRD). Locally applicable biomaterials could stabilize and foster healing of bone defects, simultaneously delivering anti-cancer compounds at systemically intolerable concentrations, overcoming drug resistance. Methods We developed silica-collagen xerogels (sicXer) and bortezomib-releasing silica-collagen xerogels (boXer) for local treatment of osteolytic bone disease and MRD. In vitro and in vivo (tissue sections) release of bortezomib was assessed by ultrahigh-performance liquid chromatography coupled to tandem mass spectrometry (UPLC-MS/MS) and time-of-flight secondary ion mass spectrometry (ToF-SIMS). Material impact on bone formation was assessed in vitro regarding osteoclast/osteoblast numbers and activity. In vivo, drilling defects in a rat- and the 5T33-myeloma mouse model were treated by both materials and assessed by immunohistochemistry, UPLC-MS/MS, µCT, and ToF-SIMS. The material’s anti-myeloma activity was assessed using ten human myeloma cell lines (HMCLs) and eight primary myeloma cell samples including four patients refractory to systemic bortezomib treatment. Results sicXer and boXer show primary stability comparable to trabecular bone. Granule size and preparation method tailor degradation as indicated by release of the xerogel components (silica and collagen) and bortezomib into culture medium. In vitro, both materials reduce osteoclast activity and do not negatively interfere with osteoblast differentiation and function. The presumed resulting net bone formation with maintained basic remodeling properties was validated in vivo in a rat bone defect model, showing significantly enhanced bone formation for boXer compared to non-treated defects. Both materials induce myeloma cell apoptosis in all HMCLs and primary myeloma cell samples. In the 5T33-myeloma mouse model, both materials stabilized drilling defects and locally controlled malignant plasma cell growth. Conclusions The combination of stabilization of fracture-prone lesions, stimulation of bone healing, and anti-tumor effect suggest clinical testing of sicXer and boXer as part of a combined systemic/local treatment strategy in multiple myeloma and non-malignant diseases.