SUPPLEMENTARY MATERIALS
METHODS
Circulating CD34+ve cell assay.
The circulating CD34+ve cell assay was performed on freshly isolated cells according to a widely
used protocol [53]. Briefly, 100 μL whole blood were incubated with 20 μL anti-CD34 monoclonal
antibody (MAb) (BD Biosciences, San Jose, CA, USA) for 20 min at room temperature in the dark.
Ammonium chloride lysing solution (3 μL) was added and incubated for 15 min to lyse red blood
cells. Cells were then centrifuged at 1500 r.p.m. for 5 min, washed in phosphate-buffered saline
(PBS) and resuspended in PBS for analysis. Acquisition was performed using flow cytometry (BD
Biosciences) and data were analyzed using Cell Quest Software (BD Biosciences). Usually, a
minimum of 60,000 events were acquired and analyzed for each sample. The number of CD34+ve
cells/μL was obtained by multiplying the percentage of CD34+ve cells by the number of white
blood cells (WBC) in 1 μL of blood.
In-vitro colony growth assays.
In-vitro colony growth assays of hematopoietic cells was performed, according to a widely used
protocol [53]. Approximately 10 ml of patients’ peripheral blood was collected in standard sterile
heparin vacutainer tubes; mononuclear cells were separated using a Ficoll-Hypaque solution
(density 1.077 g/ml; Nygard, Oslo, Norway), according to manufacturer’s protocol. Cells were
washed with PBS and pellets were resuspended in 1-2 ml of IMDM (Iscove’s Modified Dulbecco
Medium) plus 10% FBS (Fetal Bovine Serum) to count the cells obtained. Cells were then plated at
a density of 2x105 cells for 1 ml of methylcellulose medium (MethoCult™ GF H4434); cultures were
incubated at 37°C in a humidified incubator maintained at an internal atmosphere of 5% CO 2. Two
culture dishes were prepared for each patient of Group A and B. After 14 days of incubation,
culture dishes were removed and colonies were counted using an inverted microscope by two
independent observers. Counts included erythroid colonies from primitive BFU-E (burst-forming
units-Erythroid) and granulopoietic colonies from CFU-GM (colony-forming units-granulocytemacrophage). The number of colonies for each patient was determined by the arithmetic mean of
the number of colonies in the two culture dishes.
Surgical Procedure
The opening-wedge osteotomy was performed proximal to the tibial tuberosity, from the medial to
the lateral side, following a standard technique, leaving intact the lateral 5% of the tibia. The
anteromedial side of the osteotomy was fixed with a Tibia Opening-Wedge Osteotomy Plate
(Puddu Plate, Arthrex, Naples, FL) with two partially threaded 6.5-mm proximal cancellous screws
and two 4.5-mm distal fully threaded screws. All operations were performed under radiographic
control with an image intensifier, and an osteotome was used to create the osteotomy in order to
avoid thermal necrosis at the graft site. The same surgeon performed all operations, and in all
patients the osteotomy gap was filled by a hydroxyapatite and tricalcium-phosphate bone graft
substitute (HATriC, Arthrex, Naples, FL).
Radiographic evaluation
Anteroposterior standard radiographs were taken using a standard technique (65 kV, 20 mA, a
focus-film distance of 100 cm, and an equivalent effective dose of <0.01 mSv within ± 2%).
The percentage of osseointegration of the graft at the proximal and distal osteotomy levels was
calculated by estimating the density at the bone-graft junction as described in a previous study of
Dallari et al. [12]. Briefly, both the proximal and the distal bone-graft junction lines were calculated
and combined, their sum being the total length (100%) of the interface. Bone density at the bonegraft junction was defined as the dulling of the radio-transparent interface. The percentage of
dulled interface on the total length of the interface was then representative of the percentage of the
osseointegration for each radiograph. This analysis was performed at each follow-up evaluation.
For each radiographic image, the percentage of osseointegration was calculated by two blinded
observers (on two separate occasions) and a semi-quantitative score was given by each examiner
with 0+ indicating <10% of integration; 1+, 10% to 30%; 2+, 30% to 50%; 3+, 50% to 80%; and 4+,
>80%. The final score of each radiographic image was determined by the arithmetic mean of the
values given by the two blinded observers.
CT evaluation
A computed tomography (CT) evaluation (with multislice GE Healthcare CT BrightSpeed™ Excel)
of the grafted area was performed at 2 months following surgery in all patients in order to assess
the osseointegration at the bone-graft junction The osseointegration at the bone-graft junction was
estimated both through i) a simple overall qualitative score and ii) through a quantitative measure
of bone density of the newly formed bone around the bone-graft junction. Assessment on CT scans
was performed by 2 blinded observers specifically trained in bone CT examination (B.A. and C.F.).
i) The qualitative score was based on a overall evaluation of newly formed bone around the bonegraft junction. For each CT scan, all CT slices were examined by both observers to evaluate the
presence of new bone at the bone-graft junction (i.e. “gap filling”). For each CT scan, according to
the overall gap filling observed, a score from 1 to 3 was given where 1+ indicated <30% of filling,
2+ indicated 30-50% of filling, 3+ indicated > 50% of filling of the bone-graft junction, similar to the
Dallari score [12]. The arithmetic mean of the values given by the two blinded observers indicated
the qualitative score of each CT scan.
ii) The quantitative measure of bone density of the newly formed bone at the bone-graft junction
was obtained by using the Hounsfield unit (HU) scale. HU values were determined in 4 areas (i.e.
Regions Of Interest, R.O.I.) of 25 mm2 each, both at the upper ( and ) and at the lower (’ and
’) bone-graft junctions. Areas  and ’ were located at the lateral quarter of the bone-graft
junction line and areas  and ’ at the center of the bone-graft junction line (see Figure 7) . Areas
close to the osteotomy plate and screws were not evaluated because of the potential interference
created by the metallic hardware. An average HU density for each area (, , ’, ‘) was
calculated. Cancellous bone density was also calculated in an area of 25 mm2 at the proximal tibial
metaphyseal region (Y). This R.O.I “Y” was located at a distance from the plate and the screws, at
the lateral quarter of the proximal tibial metaphysis [36, 46].
SUPPLEMENTARY TABLES
Supplementary Table 1:
WBC count, the number of circulating CD34ve+cells and the number of clonogenic
progenitors (BFU-E and CFU-GM) on the day of surgery (day 0) in Group A and Group B
Group A
day 0
Group B
3
WBC count (n° x 10 /ul)
median
(range)
39.5
(20.8-51.1)
6.6
(2.8-11.1)
median
(range)
110.1
(29-256)
6.7
(5-11)
median
(range)
85.5
(58 -122)
31
(16-107)
median
(range)
44
(17-71)
5.5
(3-46)
CD34+cells /ul
BFU-E
CFU-GM
Group A: patients receiving G-CSF for 4 consecutive days, starting on day -3 before surgery;
Group B: patients undergoing the orthopaedic surgery (HTVO) without pre-operative G-CSF;
WBC= White Blood Cell; BFU-E=burst forming units- erythroid; CFU-GM=colony forming unitsgranulocyte macrophage;
differences between Group A and Group B are statistically significant (p<0.005)
Supplementary Table 2:
Physical health cluster of the SF(Short Form)-36 health survey at different time after surgery
of Group A and B
Physical Functioning (PF)
Group A
Bodily Pain (BP)
Group B
time of
observation
mean
SE
mean
basal
30d
60d
90d
6m
1y
40
27
42
60
74
77
5.3
6.3
5.1
5.8
2.7
4.2
45
9
41
54
78
85
Group A
SE
mean
SE
mean
SE
2.8
2.5
5.8
7.2
5.6
4.2
basal
30d
60d
90d
6m
1y
28
28
43
46
65
67
3.5
4.7
4.4
5.0
2.9
4.3
37
39
57
49
66
80
4.9
9.1
7.2
4.9
8.2
7.3
Role-Physical (RP)
Group A
Group B
time of
observation
General Health Perceptions (GH)
Group B
Group A
Group B
time of
observation
mean
SE
mean
SE
time of
observation
mean
SE
mean
SE
basal
30d
60d
90d
6m
1y
30
9
16
43
78
85
9.1
3.6
3.6
4.7
5.1
4.1
31
0
9
25
66
78
8.4
0
3.7
6.7
10.2
6.1
basal
30d
60d
90d
6m
1y
69
61
67
66
71
71
5.1
5.6
5.6
6.6
5.4
5.2
69
61
66
74
75
74
5.4
6.9
6.6
4.3
4.9
5.7
Group A: patients receiving G-CSF for 4 consecutive days, starting on day -3 before surgery;
Group B: patients undergoing the orthopaedic surgery (HTVO) without pre-operative G-CSF;
Time of observation: Time points of clinical assessment;
basal: preoperative clinical evaluation at the time of study start.
Abbreviations: d= days, m= months; Y=year; SE = Standard Error
Group A patients had a statistically significant improvement in PF and BP scales at days 60 and 90
compared to their basal values, while, in Group B patients, significant improvements in PF and BP
scales compared to basal values were reached no earlier than 6 months following surgery. In all
four scales no significant differences between Group A and B were noticed in the preoperative, as
well as in the end-term assessment
Supplementary Table 3:
Lysholm Knee Scale values at different time after surgery of Group A and B
Lysholm knee scale
Group A
Group B
time of
observation
mean
SE
mean
SE
basal
43
3.9
47
5.9
30 d
49
5.5
45
3.2
60d
65
4.9
58
2.9
90d
77
3.4
62
3.2
6m
87
2.1
70
4.2
1y
92
0.8
90
0.8
Group A: patients receiving G-CSF for 4 consecutive days, starting on day -3 before surgery;
Group B: patients undergoing the orthopaedic surgery (HTVO) without pre-operative G-CSF;
Time of observation: Time points of clinical assessment;
basal: preoperative clinical evaluation at the time of study start.
Abbreviations: d= days, m= months; Y=year; SE = Standard Error
p<0.05 at the two-way repeated-measure ANOVA test.
In Group A, significant improvements from baseline values were noticed at just 2 months after
surgery, whereas in Group B patients significant improvements were only recorded starting from
month 3.
Supplementary Table 4:
Semiquantitative radiographic score at different time after surgery of Group A and B
Semiquantitative Radiograhic Score
time of
observation
Group A
Group B
mean
SE
mean
SE
30 d
1.1
0.11
0.6
0.07
60d
1.9
0.1
1.3
0.1
90d
2.9
0.1
2.3
0.14
6m
3.9
0.04
3.3
0.14
4
0
4
0
1y
Group A: patients receiving G-CSF for 4 consecutive days, starting on day -3 before surgery;
Group B: patients undergoing the orthopaedic surgery (HTVO) without pre-operative G-CSF;
Time of observation: Time points of clinical assessment;
basal: preoperative clinical evaluation at the time of study start.
Abbreviations: d= days, m= months; Y=year; SE = Standard Error
p<0.001 at the two-way repeated-measures ANOVA test
Supplementary Table 5:
Bone density values of the newly formed bone in the Region of Interest at the upper and
lower bone-graft junctions
Bone Density Values (Hounsfield Unit)
Region of Interest (R.O.I.)
Group A
Group B


'
'
Y
mean
162.3
120.2
204.1
274.9
187.6
SE
19.8
23
33
35
17.7
mean
191.9
165
233.9
216
164.6
SE
32.1
20.5
34.1
18.5
16.4
p (A vs B)
0.47
0.24
0.54
0.19
0.38
Group A: patients receiving G-CSF for 4 consecutive days, starting on day -3 before surgery;
Group B: patients undergoing the orthopaedic surgery (HTVO) without pre-operative G-CSF;
Region of Interest (R.O.I.) = areas (of 25 mm2 each) at the bone-graft junctions where the bone
density of the newly formed bone was determined by Hounsfield Unit; was located at the lateral
quarter of the upper bone-graft junction;  was located at center of the upper bone-graft junction; ’
was located at the lateral quarter of the lower bone-graft junction; ’ was located at center of the
lower bone-graft junction;
Y represented the density of the cancellous bone at the lateral quarter of the proximal tibial
metaphysis.
All measure were performed at 2 months after surgery.
Abbreviations:
SE = Standard Error;
p(A vs B) = p value obtained comparing Group A values to Group B values for each R.O.I.
No statistically significant difference in the density of the newly formed bone around the bone-graft
junction was observed between the two groups. In addition, no significant differences in bone
mineral density were seen between the two patient groups at the proximal metaphiseal cancellous
bone (Y).