All experimental procedures on the New Zealand White rabbits were first
approved by the Institutional Animal Care and Use Committee of Rhode Island
Hospital.
Alzet Osmotic Pump Background and Preparation
The Alzet osmotic pump (DURECT, Cupertino, California) is an implantable
continuous infusion device that utilizes osmotic gradients to continuously
expel a specific reagent from within its loading chamber. It has been applied
in numerous animal
experiments17,18.
A variety of capacities and verified pumping rates are available under the
assumption of a normal host mammalian body temperature. In the present study,
pump model 2001D (a capacity of 200 µL and a pump rate of 8 µL/h for
twenty-four hours) and model 2004 (a capacity of 200 µL and a pump rate of
0.25 µL/h for four weeks) were used. All pumps were loaded with reagent
under sterile conditions and were primed prior to implantation by immersion in
a sterile saline solution at 37°C for three hours (model 2001D) or forty
hours (model 2004).
To use the osmotic pump as a focal delivery system from a distant implant
site, an attachable vinyl catheter was prepared. Under sterile conditions, a
15-cm-long vinyl catheter with a 0.69-mm inside diameter (number 7760; Alza)
was plugged at one end with bone wax (Ethicon-Johnson and Johnson, Piscataway,
New Jersey). A Steri-Strip (0.5 × 4 in; 3M, St. Paul, Minnesota) was
wrapped around the plugged end. Next, sixteen fenestrations with a diameter of
approximately 0.2 mm were made with a 4-0 reverse cutting needle
(Ethicon-Johnson and Johnson) over a 1-cm length of the vinyl catheter,
adjacent to the end with the Steri-Strip
(Fig. 1, A).
Phosphate-buffered saline solution was flushed through the vinyl catheter with
a 27-gauge needle to ensure that the plugged end was watertight and that the
fenestrations were patent.
Surgical Implantation of the Osmotic Pump
The pump and the implant procedure are modifications of a system proposed
by Baron et al.19.
Five-week-old New Zealand White rabbits were purchased from Millbrook Breeding
Labs (Amherst, Massachusetts) and were housed for one week prior to the
surgical procedure in the animal facility. The animals received National
Institutes of Health Open Formula Rabbit Ration (NIH 32) and water ad libitum.
At six weeks of age, the rabbits were premedicated with an intramuscular
injection of a long-acting narcotic buprenorphine (0.03 mg/kg), xylazine (6
mg/kg), and ketamine hydrochloride (20 mg/kg). One intravenous dose of
cefazolin (20 mg/kg) was given for prophylaxis against infection. The rabbits
were then intubated and placed under general anesthesia with isoflurane.
Both rabbit hind limbs were prepared with Betadine (povidone-iodine). A
16-gauge intravenous needle was passed percutaneously through the right
proximal tibial physis under fluoroscopic guidance from medial to lateral. The
needle was placed in line with the physis in the coronal plane and at the
midpoint of the anteroposterior width of the physis in the sagittal plane.
One-centimeter longitudinal skin incisions that were centered on the needle
entry and exit points were made. The attachable vinyl catheter was then passed
through the needle from lateral to medial, with the open end of the catheter
entering the point of the needle. The needle was then withdrawn from the
medial side, leaving the vinyl catheter in the physis. The vinyl catheter was
advanced medially such that the fenestrated segment rested in the physis
(Fig. 1, A and
B). The Steri-Strip tail end, now resting against the
lateral aspect of the tibia, was sutured to the iliotibial band with 3-0
Vicryl suture (polyglactin; Ethicon-Johnson and Johnson).
A 1-cm incision was made in the right inguinal region, and a subcutaneous
pocket was developed. The open end of the catheter was then tunneled
subcutaneously to this pocket and attached to the prepared osmotic pump. The
pump was then placed in the subcutaneous pocket. All skin incisions were
closed with buried, interrupted subcutaneous 3-0 Vicryl sutures.
A 1.0-mm Kirschner wire (Synthes USA, West Chester, Pennsylvania) was then
inserted percutaneously from medial to lateral at the midshaft of the tibia in
both legs with a 9.6-V power drill (Makita USA, La Mirada, California), with
care being taken to bury the end in the bone
(Fig. 1, C).
Postoperative analgesia with buprenorphine (0.03 mg/kg) was given twice a day,
every twelve hours, during the first three postoperative days.
Methylene Blue Pilot Study to Verify Pump Function
A trial methylene blue infusion study was performed to verify the
distribution of reagent throughout the physis by the osmotic pump. The
proximal part of the right tibia in four six-week-old New Zealand White
rabbits was implanted with an Alzet osmotic pump (model 2001D) prepared with
1.5% methylene blue and 30% Hypaque (diatrizoate meglumine; Amer-sham Health,
Princeton, New Jersey), following the surgical procedure described above. A
higher pumping rate was used in this experiment to expedite the evaluation of
the diffusion characteristics of the system in the physis. In addition, no
Kirschner wire was implanted since growth measurements were not necessary in
this group. The rabbits were then killed at two, four, eight, and twenty-four
hours after surgery. The proximal part of the tibiae was immediately dissected
from the rabbits, halved sagittally with a band saw, and placed in frozen
section with tissue-freezing medium (TFM; Triangle Biomedical Sciences,
Durham, North Carolina). The cut surfaces were examined under a dissecting
microscope for methylene blue staining.
Digital images were obtained with a stereomicroscope (MZ6; Leica,
Bannockburn, Illinois). The Adobe Photoshop 7.0 program (Adobe Systems, San
Jose, California) was used to obtain a more accurate determination of
diffusion distance through the use of color gradient selection. The normal
physeal cartilage color range was selected within a tolerance level of 45. The
pixels along the physis not selected then represented all areas stained by
methylene blue. The distance from the catheter center was then measured in
both the anterior and posterior directions along each sagittal section, and
these distances were averaged.
Evaluation of SDF-1 Effect on In Vivo Physeal Function
The right proximal tibial physis of twenty New Zealand White rabbits was
implanted with the fenestrated catheter attached to the Alzet osmotic pump
(model 2004). The SDF-1-treated group consisted of ten rabbits that received
an osmotic pump loaded with human recombinant SDF-1a (PeproTech, Rocky
Hill, New Jersey) at a concentration of 250 µg/mL with 30% Hypaque, in
phosphate-buffered saline solution. The Hypaque served as a radiopaque marker
of pump reservoir level. The right leg of these rabbits received the implant
(SDF-1-treated), while the left leg without the implant served as a control
(SDF-1 control). The sham-treated group consisted of ten rabbits implanted
with a pump housing phosphate-buffered saline solution and 30% Hypaque.
Similar to the SDF-1-treated group, the right leg of this group received the
sham implant (sham-treated), while the left leg received no implant (sham
control).
Serial radiographs of the tibiae were obtained every two weeks for a total
of eight weeks from the date of surgery. For the radiographs, the rabbits were
sedated with an intramuscular injection of xylazine (6 mg/kg) and ketamine (20
mg/kg). Each rabbit was placed prone on an x-ray cassette labeled with
radiopaque distance reference markers. The hind legs were fixed in neutral
rotation and full extension with tape. The xray source was positioned 90 cm
above the cassette. The radiographs were examined on a light board by an
observer blinded to the treatment. The distance (Dwire) from the
proximal tibial plateau to the proximal edge of the wire marker was measured
along the central axis of the tibia. The distance (Dcath) from the
plateau to the fenestrated catheter was also measured to assess catheter
position.
Two rabbits were randomly chosen from both the SDF-1-treated group and the
sham-treated group at two and four weeks after implantation. These rabbits
were killed with a sodium pentobarbital overdose (100 mg/kg). The remaining
rabbits from each group were killed at eight weeks with use of the same
method. The proximal part of the tibiae from the rabbits was harvested and
placed in 4% paraformaldehyde for forty-eight hours.
Qualitative and Quantitative Histology
The proximal tibial specimens were decalcified in Richman-Gelfand-Hill
solution for sixteen hours, washed with tap water for eight hours, dehydrated
with serial ethanol and xylene washes, and then were embedded in paraffin. The
samples were sectioned at three evenly distributed points along the
medial-lateral extent of the tibia. Multiple 6-µm slices were obtained from
each cut surface and deparaffinized. The sections were then stained with
hematoxylin and eosin as well as safranin O.
A histologic definition of physeal closure was used since closure at the
cellular level may precede radiographic closure by
weeks20. A
functioning physis was defined as one with intact chondrocyte
columns21. Physis
closure was conversely defined as the disruption of the columnar architecture,
invasion of the physis by vessels, and replacement of the cartilage columns by
bone21.
Physeal height was measured for the two, four, and eight-week time-points
with use of Adobe Photoshop with digital photomicrographs captured through the
SPOT digital imaging camera (Diagnostic Instruments, Sterling Heights,
Michigan). The resting zone (R) and proliferative zone (P) heights were
measured together and designated as height (R+P) since the transition between
these two zones was often indistinct, even in uninvolved, control samples. The
hyper-trophic zone (H) was measured from the first chondrocyte with a height
of >10 µm to the last intact lacuna with a hyper-trophic cell, as
previously
described21. Total
physeal height (TPH) was defined as the distance from the start of the resting
zone to the last intact lacuna along a column and was estimated as the sum of
the R+P and H measures. Most measurements were made along the longitudinal
axis of the physeal columns. For samples in which column morphology had been
disrupted, the total physeal height was measured directly along a longitudinal
axis orthogonal to the proximal and distal borders of the physis. Two slides
from the central sagittal section of each leg were examined, and four evenly
distributed areas from each slide were measured. The measurements were then
averaged. Two involved and two uninvolved proximal tibial specimens from
rabbits used in the methylene blue diffusion experiment were examined to
obtain values for physeal height at time zero (immediately after surgery).
Statistical Methods
The total growth from the proximal tibial physis was calculated at each
time-point by subtracting the immediate postoperative Dwire from
the Dwire at that time-point. These normalized
values were then averaged for each time-point. A standard paired two-tailed t
test was used to evaluate the differences in the total average growth from the
surgically treated leg and the nonsurgically treated leg within the same
rabbit. Unpaired two-tailed t tests were performed to evaluate the growth
difference between the sham-treated leg and the SDF-1-treated leg at eight
weeks of age. Standard unpaired two-tailed t tests were also used in subgroup
analysis of growth differences at eight weeks.
Repeated-measures mixed linear modeling with SAS software (version 9.1; SAS
Institute, Cary, North Carolina), with the parameters PROC MIXED, variance
components covariance structure, and REML fit, was used to test the effect of
SDF-1 on physeal height over time. The four factors used in the model were:
(1) treatment group (SDF-1 compared with sham), (2) the time after surgery
(two compared with four compared with eight weeks), (3) extremity (the right,
involved limb compared with the left, uninvolved limb), and (4) zone (R+P
compared with H). The effect of each factor on total physeal height and
individual zone heights was examined, as well as the effect of various
interactions between factors. Simple effects of significant interactions were
examined further with use of orthogonal contrasts. Least-squares calculations
were used to obtain estimates of differences between comparison groups.
All values are expressed as the mean and the standard deviation.
Significance was established with p < 0.05. All statistics were performed
with use of SAS software (version 9.1).
Methylene Blue Infusion Study
To test the ability of our modified constant infusion system to deliver
reagent throughout the physis, we performed a methylene blue diffusion study.
Sections of the proximal tibial physis with the high-output pump showed mild
staining by methylene blue at two hours and complete staining of the physis by
four hours (Fig. 2, A and
B). Qualitative staining intensity increased thereafter
to the twenty-four-hour time-point (Fig. 2,
C). The average staining distance (X), however, reached a
plateau of 3.2 ± 0.3 mm by four hours as the total diffusion distance
was limited by the diameter of the physis. Thus, the constant infusion system
was able to deliver a reagent throughout the physis.
Polyethylene Tube Position and Pump Emptying
The average radiographic migration distance (Dcath) of the
catheters was 7.3 ± 1.6 mm in the SDF-1-treated rabbits and 10.3
± 0.5 mm in the sham-treated group; the difference was significant (p =
0.001). The result is consistent with the finding that the growth of the
SDF-1-treated physis was inhibited. To determine whether final pump migration
distance limited the degree of growth inhibition achieved through the
diffusion of SDF-1, statistical testing was performed by dividing the
SDF-1-treated rabbits into two subgroups: those with Dcath of =3
mm and those with Dcath of >3 mm. Comparison of the subgroups
with respect to right proximal tibial growth at eight weeks showed no
significant difference (0.2 ± 0.3 mm; p = 0.41). The analysis was
repeated for the sham-treated group, and again there was no significant
difference (0.1 ± 0.3 mm; p = 0.60).
By four weeks, the Hypaque visible on immediate postoperative radiographs
had cleared in the remaining sixteen implanted pumps. The disappearance of the
Hypaque suggested complete expulsion of the pump contents over the four-week
course and verified the delivery rate of the pump.
Effect of SDF-1 on the Proximal Tibial Physis in Rabbits
Radiographic Analysis
The average proximal tibial growth for the SDF-1-treated and sham-treated
groups at zero, two, four, six, and eight weeks is detailed in
Figure 3. The SDF-1-treated and
sham-treated extremities began to diverge in total growth after four weeks,
after the osmotic pump had emptied. At the eight-week time-point, the growth
of the SDF-treated extremity was an average of 4.5 ± 3.0 mm (p = 0.007)
less than that of the sham-treated extremity
(Table I). Comparison of the
growth in the sham-treated and sham-control legs showed no significant
difference, with an average value of 0.2 ± 2.9 mm (p = 0.465). Thus,
pump implantation by itself did not cause a measurable growth disturbance.
In the SDF-1-treated group, the SDF-1-treated leg demonstrated a
significant difference of 3.7 ± 3.1 mm in growth compared with the
control leg (p = 0.006) (Table
I). The difference was, however, less than that observed when
SDF-1-treated legs and sham-treated legs were compared. A closer examination
of the growth curve for the control leg in the SDF-1-treated group showed a
deceleration in growth rate compared with the control leg in the sham-treated
group, beginning at the four-week time-point
(Fig. 3). The average
difference at six weeks (2.3 ± 2.5 mm) was significant (p = 0.042), but
it had resolved by the eighth week such that no significant difference in
length remained (1.0 ± 2.6 mm; p = 0.371).
Histomorphometric Analysis
Total physeal height and zonal heights for the involved and uninvolved
extremities of the sham-treated and SDF-1-treated groups at two, four, and
eight weeks after surgery are depicted in
Figure 4. The total physeal
height, as well as the individual zonal heights (resting and proliferative
zone and hyper-trophic zone), was found to decrease with time for all groups
and extremities from an average starting height of 673 ± 72 µm at
time zero (p < 0.0001). Orthogonal contrasts at each time-point showed no
significant differences in total physeal or individual zone heights between
all extremities at two weeks (p = 0.5529) and four weeks (p = 0.9566).
However, for the SDF-1-treated leg at eight weeks, the total physeal height
(225 ± 70 µm) was significantly less than that of the control leg in
the SDF-1-treated group (406 ± 50 µm), the sham-treated leg (399
± 66 µm), and the control leg in the sham-treated group (373
± 54 µm) (p < 0.0001). Reliable measurements of the individual
zone heights at the eight-week time-point for the SDF-1-treated physes could
not be obtained since a distinct border was absent between the proliferative
and hypertrophic zones.
Qualitative Histologic Analysis
Marked differences in physeal morphology between the SDF-1-treated and
sham-treated groups became apparent at the eight-week time-point. While the
sham-treated physes demonstrated a normal columnar cellular arrangement
(Fig. 5, A and
B), the SDF-1-treated physes were narrowed with gross
disruption of the columnar organization when examined at a location remote to
the catheter insertion site (Fig. 5,
E and F). The border between the residual
proliferative and hyper-trophic zones of the SDF-1-treated physes was
indistinct, as cells were distributed haphazardly and vascular channels
extended from the metaphysis across both zone remnants. Cartilage columns in
the SDF-1-treated physes were replaced by vessels and new bone formation.
Furthermore, the lacy spongiosa found in the sham-treated physis was markedly
decreased in the SDF-1-treated physis. The histologic findings were suggestive
of early physeal closure.
The intensity of the red safranin-O staining is correlated with
proteoglycan
content22.
Comparison of the sham-treated and SDF-1-treated extremities demonstrated no
visible differences at the two and four-week time-points (data not shown). At
the eight-week time-point, the sham-treated physis displayed deep safranin-O
staining in the resting, proliferative, and hypertrophic zones with a mild
decrease in staining toward the zone of provisional calcification
(Fig. 5, C and
D). As cartilage matrix with abundant proteoglycan is
calcified and then replaced by trabecular bone, safranin-O staining declines
from the hypertrophic zone to the zone of calcification. In contrast, the
safranin-O staining of the SDF-1-treated physis remained homogeneous in the
disorganized proliferative zone remnant. Loss of safranin-O staining was
observed at the distal half of the remnant hypertrophic zone, in areas of new
calcification, and along the periphery of the invading vasculature
(Fig. 5, G and
H). Distal to the hypertrophic zone, proteoglycan
remnants lay in beds of newly formed lamellar bone.
Comparison of the physeal morphology among the control legs in the
SDF-1-treated group, the sham-treated legs, and the control legs in the
sham-treated group at the two, four, and eight-week time-points demonstrated
the previously described physeal narrowing with time from surgery, but there
were no differences in cellular organization or proteoglycan content with
hematoxylin and eosin or safranin-O staining (data not shown).
Effect of Intraphyseal SDF-1 Infusion on Surrounding Articular
Cartilage According to Histologic Analysis
Qualitative examination of the tibial articular cartilage harvested from
the SDF-1-treated and sham-treated legs at eight weeks demonstrated none of
the hallmark histologic signs of
arthritis23. The
articular cartilage of the SDF-1-treated legs exhibited no surface
irregularities or clefts, normal chondrocyte organization and number, and
robust staining of the matrix with safranin O, suggesting intact proteoglycan
content (Fig. 6). Because the
synovium and other intra-articular tissues were not preserved in the sample
preparation, they were not examined.
The mechanism of physeal closure continues to be poorly understood. It is
known that mechanical and biochemical factors influence physeal
function24. Up to
the present, methods of epiphysiodesis have focused on mechanical ablation of
the
physis25-27.
In this study, we showed that application of a chemokine SDF-1 in high
concentrations with a continuous infusion system causes thinning and early
closure of the proximal tibial physis of rabbits. This demonstrates that
epiphysiodesis can be achieved biologically by applying an agent such as SDF-1
with use of a local delivery system.
The study of physeal function in vivo has typically relied on systemic
administration of agents. Few investigators have used targeted delivery
methods, and most of them used serial injections into the joint, into the
osseous metaphysis or epiphysis, into the periarticular arteries, or directly
into the
physis28-31.
A continuous, implantable, direct delivery system offers a practical model for
studying the local effects of various reagents on the physis and a therapeutic
model by which agents for use in epiphysiodesis may be evaluated. Baron et al.
developed a constant infusion system using the Alzet osmotic pump connected to
a small-gauge needle inserted several millimeters into the proximal tibial
physis of
rabbits19,32,33.
Although the system was reported to distribute the compound throughout the
physis, no direct evidence was provided to confirm this effect. Furthermore,
diffusion of the reagent seemed to be limited by the short length of the
implanted needle and by the presence of only a single delivery channel.
As a component of the present study, a modification of the approach of
Baron et al. was developed to permit a theoretically more rapid and even
distribution of delivered re-agent. A fenestrated catheter was inserted
percutaneously under fluoroscopic guidance through the width of the physis and
was connected to the Alzet osmotic pump. Comparison of the proximal tibial
growth of the sham-treated rabbit with its uninvolved, control leg in this
experiment demonstrated no deleterious growth effects from the implantation.
Furthermore, the methylene blue staining study with use of a high-output pump
confirmed the distribution of the delivered agent throughout the physis by
four hours. The physeal diffusion rate is likely peculiar to the administered
compound and is related to parameters such as the degree of calcification of
the matrix, charge of the molecule, and molecular
size34. A potential
drawback to the method is the observed migration of the catheter from the
physis in the majority of the implants, although no ultimate effect on growth
inhibition was observed in this experiment. Nevertheless, the current
technique offers a facile, verified alternative method to test the direct
action of biologic agents on mammalian physes.
In this study, we tested the effect of a chemokine SDF-1 on physeal
function in vivo. SDF-1 has garnered much interest in bone physiology for its
role as a prime osteoblast-signaling molecule and for its association with
cartilage degradation in rheumatoid arthritis and osteoarthritis. Although the
SDF-1 receptor CXCR4 has been shown to exist on hypertrophic phy-seal
chondrocytes, the role of SDF-1 in in vivo physeal function has not been
studied. The current experiment suggests that SDF-1 may have a role in physeal
closure. Targeted delivery of SDF-1 to the proximal tibial physis in New
Zealand White rabbits through a modified constant infusion system resulted in
growth inhibition. Radiographic analysis demonstrated significant shortening
of the SDF-1-treated tibiae compared with the operated tibiae from the sham
group. Histologic analysis at the eight-week time-point demonstrated physeal
narrowing and changes indicative of physeal closure.
Physeal closure occurs only after the differentiation and proliferation of
chondrocytes cannot match chondrocyte apoptosis and matrix calcification. The
histologic pattern observed in the SDF-1-treated physes resembles that of
normal and induced closure in mammalian
physes20,35-37.
Thus, SDF-1 appears to accelerate a physiologic pathway for the termination of
physis function. However, the exact mechanism for growth arrest with SDF-1 is
unclear from this experiment. The initial hypothesis of SDF-1 action in
promoting matrix degradation is supported by the advancing wave of
proteoglycan loss in the hypertrophic zone and adjacent to invading vessels of
the SDF-1-treated physis. Yet, this fails to explain the marked disruption of
the proliferative zone cell columns and the apparently homogeneous
proteoglycan content in this zone. An alternate hypothesis is that SDF-1 may
activate a tertiary messenger, such as TGF-ß, through the induction of
matrix metalloproteinases in the hypertrophic zone, which alters chondrocyte
recruitment and proliferation. To exemplify, TGF-ß, a known inhibitor of
proliferative chondrocyte differentiation to the hypertrophic phenotype, is
found in the hypertrophic zone and is activated by
MMP-1324. Since
SDF-1 induces MMP-13 expression in
chondrocytes15,
TGF-ß activation may therefore be the downstream target of SDF-1-induced
MMP expression and can explain the observed histologic abnormalities in the
proliferative zone. The hypothesis requires testing in future experiments.
The concentration of SDF-1 used in this experiment is far greater than the
physiologic levels. The SDF-1 concentration has been reported to be
approximately 70 ± 5 ng/mL in normal human synovial fluid and 750
± 80 ng/mL in rheumatoid synovial
fluid13. The
concentration of SDF-1 chosen for this experiment was approximately 3500 times
that found in normal synovial fluid. Since each chondrocyte sees an unknown,
but likely diminishing, concentration of SDF-1 along its diffusion gradient
from the catheter, and since metaphyseal and epiphy-seal vascular loops can
potentially siphon the administered agent, a high concentration was chosen to
facilitate delivery of the agent to the physis. The discrepancy in physiologic
and experimental concentrations of SDF-1 limits the interpretation of these
results for normal physeal function. Here, one can only conclude that the
physeal chondrocytes respond to nonphysio-logic doses of SDF-1 by premature
closure.
Interestingly, the radiographic growth data suggest that SDF-1 at the
infused concentration may have had transient systemic effects on the growth of
other long bones. Although SDF-1 was applied directly to the proximal part of
the right tibia, the uninvolved, left control leg of the SDF-1-treated rabbits
demonstrated mild growth inhibition beginning at four weeks compared with the
same leg of the sham-treated group. The difference became significant at six
weeks and then resolved by the eighth postoperative week. The effect can be
explained if the local infusion of SDF-1 yielded a systemic response.
Concern about the systemic effects of SDF-1 is warranted, given the
ubiquitous distribution of the SDF-1-CXCR4 axis throughout the mammalian body
and the association of articular cartilage degeneration in patients with
rheumatoid arthritis or osteoarthritis with high concentrations of SDF-1 in
synovial fluid13.
Although the histologic examination of the specimens of proximal tibial
articular cartilage from the SDF-1-treated rabbits at all time-points
demonstrated no qualitative disruption of morphology, other physes, synovial
tissue, and cartilaginous tissue were not examined in this experiment. A
necessary future experiment will involve a more complete histologic and
radiographic analysis of the cartilaginous and visceral tissues to determine
any untoward systemic effects of SDF-1.
Alternatively, the reversible growth inhibition observed may be useful as
an indicator of the minimum concentration of SDF-1 required to slow physeal
function. In the absence of serum samples from the rabbits at each time-point,
the maximum serum SDF-1 concentration achieved through complete emptying of
the osmotic pump may be estimated as 285 ng/mL (intravascular concentration =
pump concentration × pump volume/blood volume; blood volume [mL] = 0.07
× weight
[g]38; average
weight [wt] of a fourteen-week-old New Zealand White rabbit = 2500
g39). The
concentration is likely overestimated since the extracellular space, SDF-1
half-life, and potential tissue binding are not considered. However, the
calculated level is comparable with the level found in osteoarthritis and
suggests that SDF-1 at this concentration may cause physiologic slowing of the
physis14. Future
work will attempt to identify the optimal concentration of SDF-1 that will
result in physeal closure while minimizing the effects on other physes. A
system that uses differing concentrations of SDF-1 to vary the rate of growth
from a physis, possibly replacing the need for epiphysiodesis, can also be
envisioned.
In summary, SDF-1 appears to play a role in inducing physeal closure, as
its direct application in high concentrations with a continuous infusion
system caused the inhibition of physeal function in the proximal tibial physis
of rabbits. More work is required before its use in clinical applications can
be realized, since the optimal local concentration of SDF-1 to effect biologic
epiphysiodesis and limit systemic toxicity is unknown and the systemic safety
of SDF-1 in mammals has not been elucidated. However, greater understanding of
such cellular signals has the potential to offer biological therapeutic
alternatives to the treatment of limb-length inequality and other progressive
pediatric deformities. ?