Table 1.

Outcome measure and statistical analysis table.

Objective	Outcome measure	Time points of evaluation	Statistical analysis
Principal objective: to determine if the combined treatment of the ATMP, hBM-MSC + biomaterial (G2), is superior to the control, iliac crest autologous graft (G1), to obtain radiological and clinical bone consolidation	Bone consolidation is considered achieved when meeting the three following criteria: (1) Radiographic bone bridging: new bone formation across the fracture site visible in 3/4 cortices, on at least 3/4 views (2) Clinical healing: pain less than 3 in a NRS (0 to 10) for pain during full weight bearing and without weight bearing [36] (3) No further surgical major intervention (nail replacement, plate replacement, or replacement of all components of the previous surgery)	At 12 months after surgery	Superiority (target delta of 20%). Percentages will be estimated using a log-binomial regression model including the treatment, manufacturing site, and baseline value of the REBORNE scale as covariates. If unexpectedly the model does not fit, the Poisson long-link distribution function with a robust variance estimator will be used instead.
SO1: to determine if the low dose of hBM-MSC + biomaterial (G2b) is noninferior to high dose of hBM-MSC + biomaterial (G2a) to obtain radiological consolidation	Score calculated in xR images from REBORNE radiological scale following the formula: REBORNE score = ∑(internal, external, anterior, and posterior cortical value∗)/(4 × number of evaluable cortices)	At 12 months after surgery	Noninferiority (target delta of 10%). A log-transformed data of the values will be used in a mix model for repeated measurements, only if the primary outcome reaches statistical significance. Differences between arms will be estimated through adjusted means, standard errors, and its 95% CI
SO2: to compare the percentage of bone consolidation between G1 versus G2 and G2a versus G2b treatment arms	As defined in the principal objective	(i) G1/G2: at 6 and 24 months after surgery (ii) G2a/G2b: at 6, 12, and 24 months after surgery	Fisher's exact test
SO3: to compare the radiological consolidation between G1 versus G2 and G2a versus G2b treatment arms	As defined in SO1	At baseline and 6, 12, and 24 months after surgery	t-test or Mann–Whitney test
SO4: to compare pain with and without weight bearing using the G1 versus G2 and G2a versus G2b treatment arms	Pain score using the Numeric Rating Scale from 0 to 10, when 0 = no pain at all and 10 = worst pain ever	At baseline and 6, 12, and 24 months after surgery	Fisher's exact test Mann–Whitney test
SO5: to compare the rate of further surgical intervention at the callus site between G1 versus G2 and G2a versus G2b treatment arms	Further surgical intervention at the callus site is considered when nail replacement, plate replacement, or replacement of all components of the previous surgery are performed.	At 6, 12, and 24 months after surgery	Fisher's exact test
SO6: to compare the early and global complication rate between G1 versus G2 and G2a versus G2b treatment arms	Early (<3 months) and global complications included the following: (i) AE related to the product application process (surgical or other, including BM or ICA harvesting) (ii) Local heterotopic ossification (iii) Local bone resorption (iv) Local osteolysis (v) Local and general infection (vi) Vascular problems (ischemia, phlebitis) (vii) Neurological problems (viii) Unexpected events (e.g., hypersensitivity, immunological, and toxic) (ix) AE related to concomitant medication (e.g., anaesthetics)	At 6, 12, and 24 months after surgery	Fisher's exact test
SO7: to assess the safety of autologous hBM-MSCs between the G1 and G2 treatment arms	Safety understood as early or global complication rates and SAE rates, related to the use of hBM-MSCs	At 6, 12, and 24 months after surgery	Fisher's exact test
SO8: to identify the factors associated with bone regeneration between G1 versus G2 and G2a versus G2b treatment arms	Association outcomes (β, RR)	At 6, 12, and 24 months after surgery	Log-binomial regression model and mix model for repeated measurements∗∗
SO9: to compare the physical and mental health status between G1 versus G2 and G2a versus G2b treatment arms	SF-36 Health Survey score	At baseline and 6, 12, and 24 months after surgery	t-test or Mann–Whitney test

SO: secondary objective; NRS: Numeric Rating Scale; AE: adverse event. ^∗Cortical value: 1 point if fracture is unchanged, 2 points if callus is noncontinuous, 3 points if callus is continuous but fracture is still apparent, 4 points if callus is with same density as cortical, and 0 points if noninterpretable or nonvisible. ^∗∗Variables of interest: manufacturing centre (nominal), anatomical site of the fracture (femur, humerus, and tibia), sex (male, female), smoking habit (yes, no), and time since acute fracture (months).