ity to occlude these sinuses and to reliably induce infec-
tions.15,16 Histological investigations of this animal model
have demonstrated epithelial desquamation, goblet cell
hyperplasia, fibrosis, and polyposis as changes identified
in conjunction with rhinosinusitis.17 Inflammatory mark-
ers, mucosal metabolism, and ultrastructure and ciliary
structure have all been elucidated in the rabbit mod-
el.18–20 Furthermore, prior studies involving surgical in-
terventions in rabbits have demonstrated that only per-
forming a middle meatal antrostomy may be insufficient
to allow recovery from maxillary sinusitis and resolve the
extent of pathological disease evident in bacterially in-
duced (Staphylococcus aureus, Bacteroides fragilis) sinus-
itis.21–22 Such findings support the clinical observations
that additional therapies are required to improve clinical
outcomes in certain treatment situations.
Recent studies from our laboratory suggest that bone
itself may also play a significant role in experimentally
induced sinusitis.11,15,23,24 Initial experiments involving
induction of pseudomonal sinusitis infections in rabbits
demonstrated bone architectural alterations, including a
coordinated osteoclasis, appositional bone formation adja-
cent to the infected sinus followed by intramembranous
bone remodeling, as rapidly as 4 days after bacterial in-
oculation.15 Clinically, within the spectrum of acute to
chronic sinusitis, nasal endoscopy and computed tomog-
raphy (CT) studies have shown bone to undergo resorption
followed by subsequent hyperostosis. In addition, in the
postoperative patient we have noted that localized persis-
tent inflammation occurs until the underlying bone is
removed. Our previous work in the clinical environment
has also demonstrated that these histological changes are
compatible with a diagnosis of osteomyelitis.11 Using his-
tomorphometry and tetracycline labeling techniques, we
ascertained that ethmoid bone underwent rapid remodel-
ing in CRS that was histologically identical to the remod-
eling seen in osteomyelitis.23 However, as is frequently
the case in chronic osteomyelitis, we were unable to iden-
tify organisms in the underlying bone.23
Our most recent experiment demonstrated the ability
for pseudomonal sinusitis, in the presence of surgical in-
tervention, to involve bone both in proximity to and at a
distance from the site of primary infection, without the
necessity for intervening mucosal disease.24 However, this
preliminary study was performed with only one organism,
and a second procedure was performed during the study
that could, potentially, have affected the results. During
the earlier study, some bone was harvested from the side
of the infection and implanted submucosally on the non-
infected side. Histologically, this implanted bone was seen
to resorb without inducing adjacent inflammation. How-
ever, despite the fact that the changes within the bone
could be traced in a significant number of cases directly to
the experimentally infected maxillary sinus, the addi-
tional surgery also created the possibility that the inflam-
mation seen in the bone on the noninfected side might, in
some cases, result from the surgical exploration or, poten-
tially, occur as a result of the resorbed bone. Therefore,
the overall aim of the current study was to further eluci-
date the pathogenic role of bone in chronic rhinosinusitis.
In particular, the present study attempted to deter-
mine whether experimentally induced local sinusitis with
minimal bone and soft tissue damage using either Pseudo-
monas aeruginosa or Staphylococcus aureus alone can
stimulate inflammation at a distant site without involve-
ment of any intervening mucosa. S aureus accounts for a
significant percentage of aerobic pathogens in sinusitis,
and P aeruginosa has been the most frequently cultured
source of severe recalcitrant CRS infections.9,25 In addi-
tion, P aeruginosa, unlike many other acute sinusitis
pathogens, has demonstrated the ability to cause a long-
term sinusitis in the rabbit model.
MATERIALS AND METHODS
Animal Passage
Approval for all experimental procedures was obtained from
the University of Pennsylvania Institutional Animal Care and
Use Committee (IACUC). Four New Zealand White rabbits were
used for the animal passage portion of the study, two for each
bacterial strain. Both bacterial strains, P aeruginosa and S au-
reus, were obtained from nasal, pharyngeal, and sinus isolates
from the American Type Culture Collection (ATCC) (Manassas,
VA). The animals were housed at the animal husbandry facility
for at least 4 days before a surgical procedure was performed. The
rabbits were anesthetized with ketamine and xylazine, and a
pulse oximeter was attached to monitor oxygenation status. Vital
signs were recorded by the veterinary technician, and the nasal
dorsum was shaved and prepped in a sterile manner. One percent
Xylocaine was injected subcutaneously along the skin on the
nasal surface; then the animals were intubated and oxygen de-
livery was initiated and provided throughout the procedure. In-
halational anesthetic (isoflurane) was also provided throughout
the procedure to keep the rabbit anesthetized. Using a No. 15
blade, a vertical incision was made approximately 0.5 cm to the
right of the midline. The periosteal layer in the intended drilling
region was elevated, and a small periosteal flap was excised using
a scalpel. The excised periosteal tissue was placed in warm saline.
With the aid of a temporal bone drill, a small opening was drilled
in the maxillary bone by saucerizing a small area until a window
of maxillary mucosa was visualized. This remaining mucosal
layer was incised with a scalpel, and an opening approximately 5
mm wide into the maxillary sinus was exposed. Using a rongeur, a
segment of bone located superolateral to the sinusotomy site was
carefully removed to allow increased exposure of the maxillary os-
tium. To adequately occlude the semilunar maxillary sinus ostium,
the mucosal surface of the ostium was abraded and a small cotton
pledget was cut to size and inserted into the ostium with the aid of
an alligator instrument. The periosteal flap was retrieved from the
saline solution and laid over the drilling site to further ensure that
complete ostium occlusion had been achieved. Valsalva maneuver
was then performed by the anesthetist, and the periosteal flap was
observed for any movement resulting from the inevitable air leakage
around the uncuffed endotracheal tube. Any movement of the peri-
osteum placed over the sinus opening required removal and reposi-
tioning of the cotton pledget and repetition of the flap procedure.
Once the flap was immobile, the pledget was lightly moistened with
cyanomethacrylate glue gel to ensure maintenance of the cotton
ostial plug in the appropriate position. Next, 0.4 mL of the bacterial
inoculum was carefully instilled into the maxillary sinus using a
25-gauge needle. The skin was sutured with a running 2.0 Vicryl
suture to close, and the rabbits were provided with buprenorphine
HCl (0.02–0.1 mg/kg intramuscularly) every 6 hours for 24 hours
after surgery for analgesia.
All four animals were sacrificed 4 days after surgery with a
sedating ketamine dose followed by a sodium pentobarbital solu-
Laryngoscope 112: November 2002
1952
Khalid et al.: Chronic Rhinosinusitis