Methods for biological reconstruction for bone defect after malignant bone or soft tissue tumor resection can include recycled autograft (such as frozen, pasteurized, and extracorporeally radiated autograft [[1], [2], [3], [4], [5], [6], [7]]), vascularized autograft [6,[8], [9], [10]], allograft [6,[11], [12], [13], [14]], or bone transport [15,16]. However, the high incidence of complications, such as non-union and fracture, is a problem for both patients and surgeons. Among these complications, postoperative infection results in the worst oncological and functional outcomes [11]. Infection rates are significantly higher than in conventional orthopedic surgeries. For example, they reach up to 32% in allograft [6,12], 9.1–22.2% in vascularized autograft [6,8,9], 5.9–15% in frozen autograft [1,2], 13% in pasteurized autograft [3], 12.5–26.3% in extracorporeally radiated autograft [17] and 10.0–18.1 % in bone transport [15,16].

To control and reduce postoperative infection rates, we need a comprehensive understanding of the etiology of infection with biological reconstruction in malignant bone or soft tissue tumors regarding incidence and risk factors. However, due to the need for considerable complex surgical techniques and a lower frequency of diaphyseal tumors, the application of biological reconstruction seems less common than that of prosthesis reconstruction, making it more challenging to gather the necessary data.

Several practical options for data accumulation exist, including systematic reviews, prospective or retrospective multicenter and high-volume center surveys, and analysis of registry data. Multicenter/high-volume center analysis has several advantages, such as being based on a detailed study design, including strict inclusion criteria, defined outcomes, and measurement of independent variables. On the other hand, the time taken to amass a suitable sample size is a striking disadvantage compared with systematic reviews [6,17] or analyses of registry data [13], in which large samples can be more easily obtained [3,11,18,19].

To our knowledge, no study has used registry data to estimate postoperative infection incidence and risk factors in malignant bone and soft tissue tumor resection with biological reconstruction. In this study, we collated data from a large number of patients through the Bone and Soft Tissue Tumor (BSTT) Registry in Japan, launched in the 1950s by the Japanese Orthopaedic Association (JOA). All JOA-certified hospitals for musculoskeletal oncology (N = 89) are obliged to participate in this registry [20]. The registry has some disadvantages, such as the lack of several patient-related (e.g., body mass index, comorbidity, hemoglobin, and albumin), surgery-related (e.g., administration status of prophylactic antibiotics, blood loss, surgery duration), and treatment-related factors (e.g., intensive care unit control, blood transfusion) that are relevant to infection control. Nonetheless, the database does contain information on a large amount of tumor-related clinicopathological factors [20,21]. We aimed to use the database to extract the exact incidence of postoperative infection, to compare incidence among cases with different kinds of biological reconstruction modality, and to extract the risk factors for postoperative infection in the cases with biological reconstruction.