Microbiology of the ethmoid sinus following endoscopic sinus surgery. (Original Article).

Abstract

We prospectively studied the native microbiology of the ethmoid sinus following endoscopic sinus surgery in 113 patients (mean age: 41.3 yr). After each patient had regained complete mucosalization (minimum follow-up: 6 wk), we obtained a bacterial swab of the ethmoid sinus and submitted it for culture and sensitivity analysis. We then compiled data on sensitivity patterns and the number and type of bacteria isolated. Of the 113 patients, 67 (59.3%) had positive cultures, 26 (23.0%) had sterile cultures, and 20 (17.7%) had normal flora. Multiple bacteria were isolated in 31 of the 113 patients (27.4%). The most common isolates were gram-positive cocci, and the most common bacteria were staphylococcal species. A significant degree (42.9%) of beta-lactamase resistance was exhibited. We conclude that the ethmoid sinus is not microbiologically sterile following endoscopic sinus surgery.

Introduction

The advent of endoscopic sinus surgery has given otolaryngologists a minimally invasive, highly effective treatment modality for medically refractory chronic rhinosinusitis. However, not all patients respond equally well, and 6 to 8% of them will have recurrent or persistent disease that requires revision surgery. (1,2) Others will experience acute recurrences of chronic rhinosinusitis despite patent sinus ostia. (3)

In its normal state, the maxillary sinus proper is considered to be sterile, as has been shown by antral puncture studies in healthy subjects. (4) However, very little is known about the bacteriology of the maxillary and ethmoid sinuses following endoscopic sinus surgery. It has been thought possible that these sinuses become sterile following wide ventilation. On the other hand, it has also been thought possible that they acquire a bacteriologic flora similar to those of the nasal cavity. If they do, knowledge of this information would be important in helping the otolaryngologist interpret culture results in patients with recurrent acute rhinosinusitis following endoscopic sinus surgery. Identification of the typical flora in the postoperative ethmoid sinus cavity, as well as a determination of the type of bacteria expected to be colonizers rather than pathogens, would also assist the physician in selecting an appropriate antibiotic. (3)

In this article, we describe our study of the microbiology of the ethmoid sinus following endoscopic sinus surgery. Our goal was to determine whether bacterial colonization of the sinuses does in fact occur.

Patients and methods

We prospectively studied a consecutive series of 113 patients--65 women and 48 men (mean age: 41.3 yr)--who underwent endoscopic sinus surgery for chronic rhinosinusitis. All patients met the American Academy of Otolaryngology--Head and Neck Surgery's criteria for a clinical diagnosis of chronic rhinosinusitis, (5) and all had failed maximal medical management. Endoscopic sinus surgery was performed in the standard fashion, and patients were placed on a 7- to 10-day course of a first-generation cephalosporin (or a macrolide antibiotic in cases of penicillin allergy) for the immediate postoperative period. Patients were followed up in the usual manner.

At the 6-week postoperative follow-up visit, each patient was examined to ascertain if he or she met the criteria necessary to contribute an appropriate swab specimen from the ethmoid labyrinth for culture. To meet the criteria, each patient had to exhibit complete mucosalization of the ethmoid cavity and each had to have refrained from taking any antibiotic for any reason during the previous 4 weeks. Patients who did not meet these criteria were re-evaluated at subsequent visits.

Culture specimens were obtained with strict adherence to proper technique to avoid contamination from the nasal vestibule or the anterior nasal cavity. A nasal endoscope was placed into the nose, and the nasal ala was held laterally. Under endoscopic visualization, a standard microtip culture swab was placed through the nasal cavity and directly into the ethmoid labyrinth. No contact was allowed between the swab and the nasal vestibular skin, septum, and lateral nasal wall. When such contact did occur, the culture swab was discarded and a new culture was drawn. Culture specimens were drawn from the side of the nose that allowed for more direct access to the ethmoid labyrinth and a lesser chance of contact between the swab and nasal structures; 69 were drawn from the left side and 44 from the right.

Each properly obtained swab was submitted for standard aerobic culture analysis; antibiotic sensitivities were also determined when appropriate. Sensitivity analysis was conducted by using published cutoff points for identifying moderate and complete resistance to antibiotics. (6) Given that the anterior ethmoid labyrinth was open to the nasal cavity following surgery, anaerobic cultures were not obtained.

Culture and sensitivity data were tabulated and analyzed by the SPSS statistical package system version 10.0. The prevalence of recovered bacteria was then analyzed and comparisons were made among the different bacteria, Gram's staining characteristics, and bacterial sensitivities.

Results

All 113 patients were eventually able to submit an appropriate culture specimen (mean time from surgery to culture: 15.9 wk). Of this group, 26 patients (23.0%) had sterile cultures and 20 (17.7%) exhibited normal flora. In the remaining 67 patients (59.3%), culture identified 108 various isolates (table 1), most of them gram-positive cocci (table 2). Of these 67 patients, 36 had only one type of bacteria, 24 had two, and seven had three or more. None of the patients who exhibited potentially pathogenic bacteria in their postoperative culture developed a purulent sinus infection within the subsequent 2 weeks.

The sensitivity analysis of several representative anti-microbial agents revealed high degrees of resistance to penicillin, ampicillin, and erythromycin (table 3). Of all the isolates that exhibited resistance, 40.9% demonstrated multidrug resistance ([greater than or equal to]2 antibiotic resistances per isolate).

Discussion

The microbiology of chronic rhinosinusitis has been the subject of numerous studies over the past several decades. (7-11) Controversy has surrounded issues regarding the involvement of anaerobic bacteria, the causative role of organisms generally considered to be nonpathogens, and the differences between tissue cultures and mucosal surface cultures. Antral puncture studies in normal volunteers have shown that the maxillary sinus is sterile in the absence of sinonasal disease. (4)

No similar data are available on the ethmoid sinus in normal patients, but sterility has generally been inferred. Based on the presumed sterility of these sinuses, the detection of pathogenic organisms in cultures of the middle meatus or antral puncture specimens would theoretically serve to confirm a diagnosis of bacterial sinusitis in the appropriate clinical setting and help direct treatment. But in everyday practice, empiric treatment is prescribed for acute rhinosinusitis because direct culture of the maxillary and ethmoid sinuses is difficult and invasive, and access to sinonasal endoscopy for middle meatal culture is not always available.

One of the goals of endoscopic sinus surgery is to achieve free communication between the paranasal sinuses and the nasal cavity by creating wide and patent sinus ostia. During an acute postoperative exacerbation of chronic rhinosinusitis, such patency allows mucopurulent secretions to freely drain into the nose unobstructed. Under endoscopic guidance, the otolaryngologist can easily culture these secretions and plan appropriately targeted antibiotic therapy. (3) However, because postoperative sinus cavities communicate directly with the nasal cavity, it has been thought possible that they can become colonized by normal nasal bacteria. If so, the colonization of nasal bacteria could confound the interpretation of culture results and make it more difficult for the otolaryngologist to use sinonasal swabs as a basis for assessing the patient's postoperative status. It has been shown that the nasal cavity and middle meatus contain a wide variety of natural aerobic and anaerobic flora in the asymptomatic state and that fewer than 20% of patients have a sterile middle meatus. (12) More than 75% of normal adult nasal cavities harbor potential bacterial pathogens; more than 40% of these patients have staphylococcal and streptococcal species. (13)

Our study was undertaken to identify the normal flora of the ethmoid sinuses following endoscopic sinus surgery to determine whether they remain sterile or become colonized with nasal flora. We found that they are not sterile in approximately 60% of such patients. The most common organism subtypes that we identified were grampositive cocci; Staphylococcus aureus and Staphylococcus epidermidis were found in approximately equal proportions. These findings suggest that the postoperative ethmoid sinus can be colonized with bacterial flora that are very similar to those that exist in the nasal cavity. (13) Interestingly, the relative distributions of the various bacteria are quite comparable to those reported in studies of culture results in acute postoperative exacerbations of rhinosinusitis. (3) This similarity suggests that the new flora in the postoperative sinonasal cavity might be responsible for postoperative infections. It is also possible that external factors, such as allergic inflammation and viral infe ction, provide a setting in which previously quiescent colonizing bacteria become pathogenic. It seems probable that most acute postoperative exacerbations of chronic rhinosinusitis are not caused by inoculation with external bacteria.

We were somewhat surprised to find that patients could remain asymptomatic even during colonization of bacteria that are considered to be truly pathogenic. None of our patients who had pathogenic culture results ultimately progressed to purulent sinusitis within 2 weeks after culture. This finding argues that the mere presence of pathogenic bacteria is not enough to cause symptoms; instead, the presence of bacteria at a certain minimum amount or density is required to create the infection that causes symptoms. Such minimums have long been used to delineate other infections; for example, in urinary tract infection, the minimum colony count cutoff point is [10.sup.5] bacteria per ml.

The prevalence of gram-negative rods in our study--11.7%--is somewhat lower than the prevalence that has been documented in infectious exacerbations of chronic rhinosinusitis, which can exceed 20%. (11) However, we suspect that these higher prevalence rates reflect either a trend toward gram-negative sinusitis as described by others or a shift in the natural sinonasal flora created by multiple courses of antibiotics prior to endoscopic sinus surgery.

The results of our sensitivity analysis illustrate that the bacterial flora that colonize the ethmoid cavity postoperatively are often resistant to the standard first-line antibiotics that are usually considered for acute rhinosinusitis. The high rates of complete resistance to penicillin (81.5%), ampicillin (42.9%), and erythromycin (25.9%) suggest that patients who experience acute exacerbations of chronic rhinosinusitis following endoscopic sinus surgery will likely require second-line therapy on an empiric basis. Fortunately, most of these bacteria do not exhibit multidrug resistance and most are sensitive to the second-generation cephalosporins and quinolone antibiotics. We encountered only one case of methicillin-resistant Saureus in the postoperative cavity; other authors have reported that this potentially distressing situation has occurred with greater frequency. (14)

Our data provide information on the expected flora of the ethmoid sinus following endoscopic sinus surgery, and they should be considered when interpreting the results of cultures in patients with acute exacerbations of chronic rhinosinusitis. Given that colonization with gramnegative bacteria is relatively uncommon, detecting such on endoscopic culture in acute exacerbations of chronic rhinosinusitis likely means that the patient has a true pathogenic infection, and antibiotic treatment can be considered. In contrast, it is likely that many of the grampositive cocci--especially S aureus and S epidermidis--are colonizers of the nasal cavity. When gram-positive cocci are recovered on culture during acute exacerbations of chronic rhinosinusitis, treatment decisions can be more difficult. Further study might help us determine whether bacterial colony counts are more appropriate for assessing the pathogenic impact of gram-positive cocci. Because a significant percentage of even the colonizing bacteria in the post operative sinonasal cavity are beta-lactamase producers, the initial antibiotic chosen for acute exacerbations should confer beta-lactamase stability. Given the potential for bacterial resistance and the wide variety of colonizing bacteria that we found, we believe that purulent infections that fail to respond promptly to empiric therapy should be cultured and that further treatment should be based on culture and sensitivity results.

Table 1

Distribution of bacterial species

Bacteria                                No. isolates

Staphylococci                               64
  Coagulase-negative staphylococci  36
  Staphylococcus aureus             27
  Methicillin-resistant S aureus     1

Diphtheroids                                11
Alpha-hemolytic streptococci                 8
Group G streptococci                         4
Serratia marcescens                          3

Acinetobacter spp.                           2
Hemophilus influenzae                        2
Pseudomonas spp.                             2

Corynebacterium spp.                         1
Enterobacter spp.                            1
Enterobacter aerogenes                       1
Enterobacter cloacae                         1
Enterococci                                  1
Escherichia coli                             1
Hemophilus parainfluenzae                    1
Neisseria spp.                               1
Propionibacterium acnes                      1
Pseudomonas aeruginosa                       1
Salmonella enteritidis                       1
Xanthomonas spp.                             1

No growth/oral flora                        46

Total                                      154
Table 2

Distribution of isolates according to staining characteristics

Type                      n (%)

Gram-positive cocci     77 (50.0)
Gram-negative rods      18 (11.7)
Gram-positive rods      11  (7.1)
Gram-negative cocci      1  (0.7)

No growth/oral flora    46 (29.9)
Unknown                  1  (0.6)

Total                  154 (100.0)
Table 3

Patterns of sensitivity to representative antibiotics

                                             Moderately  Completely
                                  Sensitive   resistant   resistant
Drug                                n(%)        n(%)        n(%)

Clindamycin (n = 22)              21 (95.5)                1 (4.5)
Ciprofloxacin (n = 21)            20 (95.2)                1 (4.8)
Cephalexin or cefazolin (n = 28)  23 (82.1)               5 (17.9)
Erythromycin (n = 27)             19 (70.4)   1 (3.7)     7 (25.9)
Ampicillin (n = 21)               11 (52.4)   1 (4.8)     9 (42.9)
Penicillin (n = 27)                5 (18.5)              22 (81.5)

Totals (n = 146)                  99 (67.8)   2 (1.4)    45 (30.8)

References

(1.) Ramadan HH. Surgical causes of failure in endoscopic sinus surgery. Laryngoscope 1999; 109:27-9.

(2.) Bhattacharyya N. Computed tomographic staging and the fate of the dependent sinuses in revision endoscopic sinus surgery. Arch Otolaryngol Head Neck Surg 1999;125:994-9.

(3.) Bhattacharyya N, Kepnes L.J. The microbiology of recurrent rhinosinusitis after endoscopic sinus surgery. Arch Otolaryngol Head Neck Surg 1999; 125:1117-20.

(4.) Sobin J, Engquist S, Nord CE. Bacteriology of the maxillary sinus in healthy volunteers. Scand J Infect Dis 1992:24:633-5.

(5.) Lanza DC, Kennedy DW. Adultrhinosinusitis defined. Otolaryngol Head Neck Surg 1997;117(3 Pt 2): SI-7.

(6.) Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Wayne, Pa.: National Committee for Clinical Laboratory Standards, 1997.

(7.) Ramadan HH. What is the bacteriology of chronic sinusitis in adults? Am J Otolaryngol 1995;16:303-6.

(8.) Brook I. Microbiology and management of sinusitis. J Otolaryngol 1996;25:249-56.

(9.) Jiang RS, Hsu CY, Leu JF. Bacteriology of ethmoid sinus in chronic sinusitis. Am J Rhinol 1997;11:133-7.

(10.) Liu ES, Lebowitz RA, Jacobs JB, Tierno PM. The bacteriology of chronic rhinosinusitis: Results using a novel culture device. Am J Rhinol 2000;14:101-5.

(11.) Nadel DM, Lanza DC, Kennedy DW. Endoscopically guided cultures in chronic sinusitis. Am J Rhinol 1998;12:233-41.

(12.) Klossek JM, Dubreuil L, Richet H, et al. Bacteriology of the adult middle meatus. J Laryngol Otol 1996; 110:847-9.

(13.) Gluck U, Gebbers JO. The nose as bacterial reservoir: Important differences between the vestibule and cavity. Laryngoscope 2000;110:426-8.

(14.) Jiang RS, Jang JW, Hsu CY. Post-functional endoscopic sinus surgery methicillin-resistant Staphylococcus aureus sinusitis. Am J Rhinol 1999;13:273-7.

From the Division of Otolaryngology, Brigham and Women's Hospital, Boston (Dr. Bhattacharyya), the Division of Otolaryngology, Beth Israel-Deaconess Medical Center, Boston (Dr. Gopal), and the Department of Otology and Laryngology, Harvard Medical School, Boston (Dr. Bhattacharyya and Dr. Gopal).

Reprint requests: Neil Bhattacharyya, MD, Division of Otolaryngology, 333 Longwood Ave., Boston, MA 02115. Phone: [phone]; fax: [phone]; e-mail: [e-mail]