The present retrospective clinical study was conducted at the Maastricht University Medical Center+ (the Netherlands), utilizing a patient database comprising consecutive individuals treated for tibial non-union from November 2014 to July 2023. The objective of this study was to evaluate the efficacy of the applied therapeutic strategy for addressing tibial non-unions. Specifically, we comprehensively compared the polytherapy approach with the monotherapy approach.

Recently, the Non Union Scoring System (NUSS) score was introduced to provide a comprehensive assessment of all potential risk factors contributing to the development of non-union [23, 24]. Accordingly, we utilized the NUSS score to evaluate the severity of non-unions, comparing two similar populations receiving either monotherapy or polytherapy. For assessing treatment outcomes, we employed the Radiographic Union Score for Tibial fractures (RUST) score [25].

All procedures involving human participants were in accordance with the local Bioethics Committee and with the Helsinki Declaration (as revised in 2013). Inclusion criteria were as follows: (1) meeting the Food and Drug Administration (FDA) diagnostic criteria for non-union: a fracture that persists for a minimum of 9 months without signs of healing for 3 months [26], (2) age ≥ 18 years and (3) surgical treatment for a tibial non-union in our hospital. Exclusion criteria were as follows: (1) history of bone cancer, (2) pregnancy, (3) immunosuppressive drug therapy, (4) autoimmune disease and (5) neoplasia.

The data from a total of 34 patients met the inclusion criteria and were therefore analysed. Several data were retrieved from the electronic patient records. Demographic data, the type of non-union including the NUSS score, the index treatment and any subsequent treatments were recorded. Regarding the type of intervention, several biological materials are applied in our clinic, including ABG, reaming irrigation aspirator (RIA, Johnson and Johnson, USA), iFactor (Cerapedics, Westminster, CO, USA), Cerasorb (Curasan GmbH, Kleinostheim, Germany), bone marrow aspirate concentrate (BMAC, Arthrex, Naples, FL, USA), tricalcium phosphates (TCP), bone morphogenetic protein-2 (BMP-2, Medtronic, Minneapolis, MN, USA) and Polycaprolactone-tricalcium phosphate (PCL-TCP) three-dimensional (3D)-printed cages (Osteopore, Singapore). Each of these can be categorized into three distinct categories, namely MSCs, bioactive factors and osteoconductive scaffolds. Additionally, for several patients, BioActiveGlass (BAG S53P4, BonAlive, Turku, Finland) was utilized for its antibacterial properties rather than for osteoconduction. Consequently, bioglass was not classified under any specific category. In this study, monotherapy was defined as utilizing a maximum of one category out of the three types, while polytherapy involved a minimum of two out of three categories.

Firstly, we compared the outcome of the initial non-union surgical treatments between the monotherapy group (n = 19) and polytherapy group (n = 15) using the NUSS score and RUST score. Following the criteria of the RUST score, cases with a minimum score of 10 and no further non-union surgical treatment were scored as having a successful outcome. Cases with a maximum RUST score of 9 at the last follow-up visit, or cases that underwent additional surgical interventions, were considered to have a failure outcome.

Patients who received additional surgical treatment for their non-union were recorded as well. If the time between the first procedure and the additional treatment exceeded 6 months, the additional treatments were counted and a per-treatment analysis was performed accordingly. In this manner, one patient may contribute to multiple treatment categories, for example, if the first procedure consisted of a monotherapy treatment followed by polytherapy more than 6 months later. The same standardization of success and failure outcome was used to compare monotherapy per-treatment and polytherapy per-treatment. The RUST scores at 1, 3, 6, 9, 12 and 18 months post treatment were recorded, if available.

Quantitative data are presented as either mean ± standard deviation (SD) or median (P25, P75), depending on whether the data followed a normal distribution as determined by the Shapiro–Wilk test. According to data characteristics, continuous quantitative data were compared between the two groups using Student’s t-test or the Mann–Whitney U test. Categorical variables were analysed using the chi-squared and Fisher’s exact tests. For evaluating potential imbalance of risk factors, multiple regression models were performed to adjust comparisons. Outcomes were recorded in binary format (‘success’ or ‘failure’). In the comparison between the polytherapy per-treatment group and the monotherapy per-treatment group, the two groups under examination are no longer entirely independent, as they partially share several patients. Consequently, a mixed linear model was used to analyse this non-independence to compare the treatment outcome between each group. RUST scores were examined both as continuous variables and in binary form (‘RUST score ≥ 10’ or ‘RUST score < 10’) at 1, 3, 6, 9, 12 and 18 months post treatment. We evaluated the association between treatment types and overall treatment outcome in a Cox proportional hazards regression model. We conducted binary logistic regression analysis to investigate the influence of NUSS score on the outcomes of monotherapy patients. The predictive capability of NUSS was assessed using the area under the ROC curve (AUC) with a criterion of maximizing the sum of sensitivity and specificity as the diagnostic threshold. A two-sided test was employed, and statistically significant difference was defined as P value < 0.05. Data analysis was conducted using SPSS 27 (SPSS Statistics for Windows, version 27.0.0; IBM, Armonk, NY, USA), while survival analysis was performed using GraphPad Prism 10 (Prism for Windows, version 10.0.2; GraphPad Boston, MA, USA).