Meticulous débridement of contaminated soft tissues
is regarded as the most important initial step in the management
of open tibial fractures1-4. Reported
rates of infection following severe open tibial fractures have ranged
from 5% to 50%5-12.
The efficacy of high-pressure irrigation in decreasing the bacterial
load in soft tissues has been well established in the literature13-20. The advent of pulsatile lavage
has further improved the removal of bacteria from soft
tissues16,17,20. In principle,
high-pressure pulsatile lavage provides a pulse compression phase
and an interpulse decompression phase in which recoil of the soft
tissues occurs, dislodging particulate matter and bacteria. The
popularity and effectiveness of high-pressure pulsatile lavage of
soft tissues have been extrapolated to a perceived efficacy for
the débridement of bone.
The increased use of high-pressure pulsatile lavage for fracture
débridement may result in complications15,21-24.
In an in vivo study of rabbits, Dirschl et al.
found that high-pressure pulsatile lavage resulted in visible damage
at the fracture site and in delayed healing15.
In a previous study, we examined the effects of high-pressure pulsatile
lavage on contaminated human tibiae in an in vitro model21. We found that high-pressure irrigation
resulted in macroscopic damage to bone and carried surface bacteria
into the intramedullary canal.
The optimal technique for bone débridement should remove
the maximal number of adherent bacteria yet preserve the structure
and function of bone. Low-pressure pulsatile lavage has obvious
potential advantages over high-pressure lavage in terms
of decreasing the degree of damage to bone, but it may remove adherent
bacteria less effectively. In a subsequent study, we showed that
low-pressure lavage results in significantly less damage to bone
(p < 0.001) and that it is as effective as high-pressure lavage
in removing bacteria within three hours after contamination22. However, when irrigation was delayed
for more than three hours, low-pressure lavage was ineffective in
removing bacteria.
The ability of various solutions to remove adherent bacteria
from hard surfaces has been previously reported25-29.
We hypothesized that the use of certain irrigating solutions would
improve the efficacy of low-pressure lavage when irrigation was
delayed beyond three hours.
The purposes of the current study were to determine the effect
of various irrigating solutions on the number and function of osteoblasts
and osteoclasts in vitro as well as to compare
the abilities of those solutions to remove adherent bacteria from
contaminated bone and to determine whether those abilities were
improved by the use of low-pressure pulsatile lavage.
Two separate sets of experiments were performed for this study.
In the first set (Part One), the effects of various irrigating solutions
on osteoblasts and osteoclasts were determined with cell culture.
In the second set (Part Two), the efficacy of the irrigating solutions
to remove adherent bacteria from bone, with or without low-pressure
lavage, was evaluated.
Part One
Calvarial Cell Isolation
Calvariae were harvested from three-day-old C57BI/6
mice (Charles River Laboratories, St. Constant, Quebec, Canada),
as described previously30. Under
low-power magnification, the parietal bones were exposed with sharp
dissection of the overlying skin and subcutaneous tissue. Each calvaria
was removed, minced, and resuspended in phosphate-buffered saline
solution; then it was digested with collagenase (2.5 mg/mL)
for four hours at 37°C. The resulting cells were washed, pelleted, and
seeded into T-50 vented flasks at a concentration of 1 ¥ 106 cells/flask
(Becton Dickinson, Lincoln Park, New Jersey). The calvaria-derived
bone cells were then grown for seven days in alpha-minimum essential medium
containing 10% fetal bovine serum and 1% penicillin-streptomycin
(Gibco BRL, Burlington, Ontario, Canada) prior to use. The cells
were plated on 35-mm plates at a density of 5 ¥ 104 cells/plate30.
Irrigating Solutions
The following irrigating solutions were examined: 10% and
1% ethanol, 10% and 1% povidone-iodine (Becton
Dickinson Canada, Mississauga, Ontario, Canada), 4% and
1% chlorhexidine gluconate (Becton Dickinson Canada), 10% and
1% liquid soap (Huntington Laboratories, Huntington, Indiana), 10% and
1% antimicrobial wash (50 U/L of bacitracin),
and normal saline solution (control). The soap solution was prepared
by injecting 100 mL of liquid soap into 1 L of normal saline solution.
We chose concentrations that are used clinically for the irrigation
of wounds and fractures, and we chose irrigating solutions and relative
concentrations that are comparable with those found in the literature25,31-34.
In the first set of experiments, 35-mm plates of cultured calvarial
cells at day 5 were exposed for two minutes (thirty plates), ten
minutes (thirty plates), or twenty minutes (thirty plates) to 3
mL of the higher concentration of each irrigating solution (five
experimental solutions and one control [saline] solution).
Thus, for each of the six solutions, there were five plates for
each of three time points (6 ¥ 5 ¥ 3 = 90
plates). Twenty-four hours after exposure, the cells were trypsinized
(0.5% trypsin) and were stained with trypan blue. The total
number of viable cells was quantified with light microscopy and
a hemocytometer. Since the effect of the higher concentrations was
maximal at two minutes, this time point was used for all subsequent
experiments.
In subsequent experiments, the effect of irrigating solutions
on the activity of osteoblasts and osteoclasts was evaluated by
exposing 35-mm plates (five plates for each solution) with calvarial
cells to each of six solutions at two concentrations (1% and 10%)
for two minutes at day 5. At various time points thereafter, the
number of alkaline-phosphatase-positive cells (day 21), the number
of bone nodules (day 21), and the number of tartrate-resistant acid-phosphatase-positive
cells (days 14, 15, and 16) were quantified by a blinded observer. Thus,
a total of 180 tissue-culture plates were used for this experiment
(6 solutions ¥ 2 concentrations ¥ 5 plates per
solution ¥ 3 outcome measures = 180 plates).
Alkaline-Phosphatase Assay
Osteoblasts were identified histochemically with alkaline-phosphatase
staining (Sigma Chemical, St. Louis, Missouri)30,35.
The cells were fixed with 60% citrate-buffered acetone
for forty-five seconds. The diazonium salt solution was prepared
by dissolving 12 mg of fast blue salt in 50 mL of 0.01% alkaline
naphthol AS-MX phosphate. The cells were immersed in the alkaline-dye
mixture for approximately thirty minutes and then were counterstained
with a hematoxylin solution. The osteoblasts were identified by visible
cytoplasmic staining of the precipitated azo dye at the sites of
alkaline-phosphatase activity. The proportion of alkaline-phosphatase-positive cells
in a defined area (0.25 mm2) was quantified
by light microscopy at a magnification of 100 times.
Bone-Nodule Assay
Since the primary function of the osteoblast is to form bone,
osteogenic capacity now appears to be the most reliable and unambiguous
parameter for characterizing a cell type as an osteoblast. Bellows et
al. showed that enzymatically released cells from fetal calvariae
are capable of forming bone-like tissue when grown in the presence
of ascorbic acid and organic phosphate35.
Calvarial cells were cultured in alpha-minimum essential medium,
supplemented with 1% nonessential amino acids
(Gibco BRL), 10-mM b-glycerophosphate (Sigma Chemical),
and 0.5-mM ascorbic acid (Sigma Chemical) at an initial cell concentration
of 5.0 ¥ 104 cells/dish.
The medium was exchanged every three or four days. At twenty-one
days, the cells were fixed in 10% formalin and bone nodules
were stained with 0.2% alizarin red S (Sigma Chemical). Nodules
were identified as red-staining structures and were quantified under
low-power light microscopy (magnification of twenty times)30.
Tartrate-Resistant Acid-Phosphatase Assay
Osteoclast differentiation was assessed in co-cultures of murine
calvarial cells and osteoclast precursors obtained from the bone
marrow of thirty-five-day-old Swiss Webster mice (Charles River Laboratories).
Briefly, calvarial cells were plated on Thermanox discs in twenty-four-well
tissue-culture plates (Becton Dickinson) at a density of 4 ¥ 105 cells/well
and expanded for five to seven days in alpha-minimum essential medium
supplemented with 10% fetal bovine serum and 1% penicillin.
On day 7, osteoclast precursors, obtained from the femoral bone
marrow of Swiss Webster mice, were co-cultured with the calvarial
cells at a plating density of 2 ¥ 105 cells/well.
All co-cultures were performed in phenol-red-free minimum essential
medium (Gibco BRL). The medium was changed every three days. On
day 9 of co-culture, the cells were stained for tartrate-resistant
acid-phosphatase activity with use of a commercially available assay
(assay 386; Sigma Chemical). Osteoclasts were identified
as red tartrate-resistant acid-phosphatase-positive multinucleated
cells and were quantified under low-power light microscopy (magnification
of 100 times) by a blinded observer.
Part Two
Preparation and Contamination of Canine Tibiae
Fourteen tibiae from seven dogs were excised in their entirety,
and all soft tissues except the periosteum were removed. Fifty-two
10-mm transverse cut sections from the diaphyses of the fourteen canine
tibiae (four sections per tibia) were obtained with use
of a standard handheld oscillating saw (Stryker Instruments,
Kalamazoo, Michigan) fitted with a sterile number-15 blade. The
surface of each cut section was contaminated with more than 108Staphylococcus
aureus organisms per millimeter (ATCC 29213; American Type
Culture Collection [ATCC], Rockville, Maryland)
in a biologic sterility hood.
Pulsatile Lavage
Six hours after bacterial contamination, the fifty-two contaminated
transverse sections of canine tibiae were subjected to either the
irrigating solution without low-pressure lavage (twenty-four sections) or
irrigating solution with low-pressure pulsatile lavage (twenty-four
sections), or they served as controls (four sections). A standard
battery-operated pulsatile-lavage system (Surgilav Plus
Debridement System; Stryker Instruments) with a multi-orifice
(four-hole) tip (tip 207-58; Stryker Instruments), which delivered
four streams of irrigating solution perpendicular to the surface
of the bone, was used for all experiments. The irrigating tip was held
approximately 5 cm from the surface of the bone for both
high-pressure and low-pressure irrigation procedures.
This lavage system has two settings (high and low pressure) and
produces a higher peak force than most other commercially available systems
at the high-pressure setting. Moreover, at the low-pressure setting
the system delivers 14 psi (96.6 kPa) of pressure (according to
the laboratory testing at Stryker Canada [Burlington, Ontario, Canada])
with 550 pulsations per minute. Each specimen was irrigated with
a total of 500 mL of normal saline solution. The choice of 14 psi
for low-pressure irrigation was based on three factors: (1) 14 psi
is the amount of pressure delivered by the Stryker Surgilav Plus
low-pressure setting with the standard four-hole tip, (2) there
is no agreement in the literature regarding the absolute definition
of low-pressure irrigation, and (3) we wanted our experimental protocols
to be consistent with those used previously in this field21,22.
Bacterial Cultures
After irrigation with or without low-pressure lavage, each canine
tibial specimen was placed in 5 mL of Luria broth (10 g of bacto-tryptone,
5 g of bacto-yeast extract, and 10 g of sodium chloride) with 1.5% agar
(Difco Laboratories, Detroit, Michigan). Each 10-mL test tube of
broth was incubated at 37C with constant stirring for
six hours on an electric test-tube stirrer (200 rpm). At six hours, 100
mL of supernatant from each specimen was plated on blood agar (Columbia
base agar with 5% defibrinated horse blood) with use of
a sterile loop and was incubated at 37C for twenty-four hours. Colony-forming
units of Staphylococcus aureus were counted after incubation.
Statistical Analysis
The Student t test was used for comparisons between two independent
continuous variables. Single-factor analysis of variance was used
to compare the means of more than two independent groups. A Bonferroni
correction was used for multiple comparisons. All tests were two-tailed,
and a p value of less than 0.05 was considered significant. Continuous
variables were expressed as means and standard errors
of the mean.
Part One: Effects of Irrigating Solutions on
Osteoblasts and Osteoclasts
Duration of Exposure to Irrigating Solutions
Exposure of the calvarial cells for two minutes to 10% ethanol,
10% povidone-iodine, 10% antimicrobial wash, or
4% chlorhexidine gluconate resulted in cell-density decreases
of 70% (p < 0.001), 63% (p < 0.001),
70% (p < 0.001), or 69% (p < 0.001),
respectively (Fig. 1). The cells treated with normal
saline solution and the soap solution did not significantly decrease
in number when compared with the controls (grown in minimum essential
medium) (p = 0.34 and 0.44, respectively). Cell density
continued to decline with increased exposure (ten and twenty minutes) to
the irrigating solutions (Fig. 1). Since the effect of most irrigating
solutions was near maximal after as little as two minutes of exposure
to the calvarial cells, this time-period was used for all subsequent
experiments.
Effects on Alkaline-Phosphatase Staining (Osteoblast
Number)
Alkaline-phosphatase-positive cells were used as markers for
osteoblasts. While exposure to all five experimental irrigating
solutions resulted in a decreased proportion of alkaline-phosphatase-positive
cells compared with that after exposure to the saline control (p < 0.001),
the soap solution was most effective in limiting this decline (Fig. 2). Moreover,
the 10% irrigating solutions (and the 4% chlorhexidine-gluconate solution)
decreased the proportion of alkaline-phosphatase-positive cells
more than did the 1% irrigating solutions (p < 0.01).
At the higher concentrations, irrigating solutions decreased the
percentage of alkaline-phosphatase-positive cells by as much as
99% (povidone-iodine and chlorhexidine gluconate) and as
little as 42% (soap) (Fig. 2).
Effects on Bone-Nodule Formation (Osteoblast Function)
To compare the effects of various irrigating solutions on osteoblast
activity, the amount of bone-nodule formation with each solution
was quantitated. Similar to their effects on the number of osteoblasts,
the various irrigating solutions significantly decreased bone-nodule
formation (p < 0.001) (Fig. 3). The 1% soap solution
was the only one that did not significantly decrease the number
of bone nodules when compared with the saline control (p = 0.75).
At higher concentrations, all irrigating solutions resulted in greater
inhibition of bone-nodule formation (p < 0.001) (Fig. 3).
Effects on Tartrate-Resistant Acid-Phosphatase Staining
(Osteoclast Number)
Tartrate-resistant acid phosphatase, an enzyme found in the cytoplasm
of osteoclasts, was stained to quantitate the number of osteoclasts.
All irrigating solutions decreased the number of tartrate-resistant
acid-phosphatase-positive cells when compared with the saline control
(p < 0.001) (Fig. 4). Specifically, the reduction in
the number of osteoclasts ranged from as low as 18% (1% soap solution)
to as high as 97% (10% povidone-iodine solution)
(Fig. 4).
The higher-concentration solutions (10% and 4%)
inhibited osteoclasts more than did the lower-concentration solutions
(1%).
Part Two: Efficacy of Irrigating Solutions,
with and without Pulsatile Lavage, in Removing Adherent Bacteria
Irrigating Solutions without Low-Pressure Lavage
Following a six-hour incubation period, a two-minute exposure
to each of the six irrigating solutions resulted a significant removal
of bacteria from bone (p < 0.001) (Table I). Moreover,
significant differences in the magnitude of the effect were observed
among the solutions (analysis of variance, p < 0.01). The
fewest numbers of residual bacterial colony-forming units were found
after exposure to the povidone-iodine (mean, 0.67 colony-forming
unit), chlorhexidine gluconate (mean, 0.33 colony-forming unit),
and soap solutions (mean, thirty-five colony-forming units). Alternatively,
the least effective solutions were the normal saline solution and
the antimicrobial solution.
Irrigating Solutions with Low-Pressure Lavage
The efficacy of four of the six irrigating solutions, with regard
to bacterial removal from bone, was improved when the solution was
applied to the surface of the bone under low pressure (Table I). Low-pressure
irrigation of bone with normal saline, ethanol, antimicrobial, and
soap solutions resulted in 2.9, 4.8, 18.8, and thirty-five-fold decreases,
respectively, in the number of remaining bacteria when compared
with irrigation without low-pressure pulsatile lavage. Moreover,
no bacterial growth was observed following low-pressure pulsatile
lavage with soap solution. Low-pressure irrigation with povidone-iodine and
chlorhexidine-gluconate solutions resulted in near complete removal
of all adherent bacteria to bone (mean and standard error of the
mean for both, 0.33 ± 0.33 colony-forming unit).
Study Results
With in vitro bone-nodule, alkaline-phosphatase,
and tartrate-resistant acid-phosphatase assays, we showed that irrigating
solutions resulted in both a time and a dose-dependent decrease
in calvarial cell density, that soap solution resulted in the smallest
decline in the number and function of osteoblasts, and that soap
solution resulted in the smallest decline in the number of osteoclasts.
Additionally, with an in vitro model of contaminated
canine cortical bone, we showed that low-pressure lavage
with saline solution alone results in significant (but not complete) removal
of residual bacteria adherent to bone and that the addition of soap
solution results in the complete removal of adherent bacteria from
bone following a six-hour delay.
Effects of Irrigating Solutions on Bone
The effects of povidone-iodine (Betadine) and bacitracin solutions
on cultured chick osteoblasts have previously been reported32. Kaysinger et al. found that a two-minute
exposure to 5% Betadine solution resulted in a 30% decline in
lactate production (a marker of glycolytic energy metabolism) and
a 90% decline in DNA synthesis (a marker of cell number)32. These findings support our observation
that the 1% povidone-iodine solution resulted
in decreases in calvarial cell density, osteoblast numbers,
and bone-nodule formation. Kaysinger et al. did not detect
any significant decrease in lactate production and they detected
less than a 25% decline in osteoblast DNA synthesis with
antimicrobial solutions (50,000 U/L of bacitracin)32, in contrast to the dramatic effects
of the povidone-iodine solution. Our findings suggest that antimicrobial
wash inhibited the formation of bone and decreased the number of
osteoblasts to the same degree as the povidone-iodine solution.
The differences that we observed might have been related, in part,
to the fact that we used an in vitro murine cell-culture
system, whereas Kaysinger et al. used a chick osteoblast-culture
system.
The effects of ethanol on bone metabolism have previously
been reported36,37. Klein et al.
examined the effect of ethanol on the number of osteoblasts in an
osteoblast-like osteosarcoma cell line and found a dose-dependent
inhibition of DNA synthesis37.
However, they did not find a decline in alkaline-phosphatase
activity. Chavassieux et al. evaluated the dose-dependent effects
of ethanol on human osteoblastic cells and reported a significant
reduction in alkaline-phosphatase activity36,
contrary to the findings of Klein et al. We found a significant
reduction in alkaline-phosphatase staining and a reduction in bone-nodule
formation after exposure to ethanol. These findings suggest a direct toxic
effect of ethanol on osteoblasts.
The effect of soap or other detergents on bone formation has
not been previously quantified in the literature, to our knowledge.
Tarbox et al., in an in vivo study of thirty Sprague Dawley
rats, found that benzalkonium chloride (a cationic detergent) did not
alter the histologic appearance of bone or cartilage38. However, they did not quantify
bone-formation parameters such as osteoid thickness or
osteoblast surface with well-established histomorphometric techniques.
Our results suggest that, at the concentrations tested, soap solutions
preserve osteoblast activity (bone-nodule formation) despite decreasing the
overall number of available osteoblasts (alkaline-phosphatase-positive
cells). Moreover, the finding that soap solution did not decrease
overall calvarial cell density (at the two-minute exposure) but
did decrease the number of osteoblasts is likely explained by increases
in another cell type. Osteoblasts are derived from a stem cell that
has the potential to differentiate into adipocytes, chondrocytes,
fibroblasts, and muscle cells. There have been previous reports
of an inverse relationship between adipocytes and osteoblasts39,40.
It has been well established that osteoblast and osteoclast interactions
are coupled41. We were unable
to find any previous studies that examined the effect of irrigating
solutions on the number or activity of osteoclasts. Our results
suggest that, among the irrigating solutions examined, the soap
solution best preserved osteoclast numbers. This finding is important
given the coupled interactions of bone formation and bone resorption.
Low-Pressure Irrigation of Bone
High-pressure pulsatile lavage of contaminated soft tissues has
been extensively tested13-20,31,42-44.
It is believed that the elastic recoil of soft tissues between high-pressure
pulses dislodges and removes contaminants from the wound14,18. Previous investigators have
used pressures between 1 and 75 psi (6.9 and 517.5 kPa) for the débridement
of wounds14,18.
There are emerging reports of the effects of high-pressure lavage
on bone. Dirschl et al., in a rabbit femoral model of fracture-healing,
reported that high-pressure lavage resulted in visible damage to bone,
a trend toward decreased early new-bone formation, and significantly
less viable bone at the fracture site compared with controls (p < 0.001)15. Moreover, in their study, 30% of
the osteotomy sites treated with high-pressure lavage failed to unite
compared with 20% of the sites treated with low-pressure
lavage and the sites in the control group. Furthermore, in a previous
study, we showed that high-pressure lavage (70 psi [483
kPa]) visibly damages the marrow contents to a depth 4
cm from the lavage site21. West
et al. supported these findings with electron microscopy; they demonstrated
that high-pressure lavage left vacant interstices of bone devoid
of cells23.
It has previously been reported that low-pressure pulsatile lavage
with normal saline solution is as effective as high-pressure lavage
in removing adherent bacteria from bone if the bone is debrided within
three hours after contamination22.
Removal of Adherent Bacteria
While several investigators have evaluated the efficacy of various
irrigating solutions in removing adherent bacteria from soft tissues25,28,31,33,38,42,44-47, few have examined
their efficacy on hard surfaces26,29,34.
Anglen et al. examined the ability of pressure irrigation with soap,
antimicrobial (bacitracin), and normal saline solutions to remove Staphylococcus
aureus from cortical bone fragments26.
They found pressure irrigation with soap solutions to be the most
effective method of reducing residual colony counts of bacteria;
however, the effective pressure used in their study was not reported. Moreover,
it is unclear whether they allowed a time delay before irrigating
the bone pieces. White et al. evaluated the antimicrobial activity
of 2.0% chlorhexidine rinses during root canals and reported improved
antimicrobial activity for up to seventy-two hours after rinses34. Gravett et al., in a randomized
trial of 500 consecutive wound irrigations with either 1% povidone-iodine
or normal saline solution, found that 1% povidone-iodine
solution significantly reduced the risk of wound infection42. Unfortunately, they did not report
the efficacy of the 1% povidone-iodine solution in a subgroup
of patients with exposed bone.
Limitations of the Current Study
While in vitro studies can provide important
information regarding the mechanisms of bone metabolism, it remains unclear
whether the substantially better bacterial removal by low-pressure
lavage with soap solution in the current study can be extrapolated
to grossly contaminated open fractures of the tibia. Future studies
should be performed to examine the ability of soaps (and other detergents)
to remove bacteria from bone in in vivo models
of fractures contaminated with a variety of different bacteria.
These additional studies will allow investigators to further develop
the biologic rationale for conducting clinical trials in this area.
In conclusion, while there are reports in the literature favoring
many of the irrigating solutions used in the current study, our
results suggest that soap solutions delivered under low pressure
are most effective in removing adherent bacteria and least disruptive
to osteoblasts and osteoclasts. The mechanism of soap’s
action on bone lies primarily in its ability to form micelles. Micelles
of soap (and of other such detergents) have hydrophilic (water-attracting)
and hydrophobic (water-repelling) ends. The soap micelles’ hydrophilic
ends surround bacteria and interfere with bacterial adherence to
bone.