The Role of Free Radicals in Chronic Rhinosinusitis

Objective  To determine whether there is an increased amount of free radical–mediated damage in diseased vs healthy tissue from patients with chronic rhinosinusitis.

Design  Pathophysiologic study. Samples of heathly and diseased tissue were taken from each patient. Lipid peroxides (LPOs) are a by-product of free radical–mediated damage; LPO levels and LPO/protein ratios were determined for each patient.

Subjects  Consecutive series of 13 human subjects undergoing functional endoscopic sinus surgery to treat chronic rhinosinusitis.

Results  The mean LPO/protein ratio for healthy tissue was 3.52 × 10^-5, while that for the diseased tissue was 3.49 × 10^-5. There was no statistically significant difference in the LPO/protein ratio between healthy and diseased tissue (95% confidence interval, −3.00 × 10^-5 to 2.94 × 10^-5).

Conclusion  Free radical–induced damage, if present, was the same in infected and control tissues in this pilot investigation into the pathophysiologic characteristics of human chronic rhinosinusitis.

FREE RADICALS are highly reactive species containing 1 or more unpaired electrons. They are produced in vivo during normal metabolism by enzymes such as xanthine oxidase and nitric oxide synthase and most abundantly by the electron transport chain during oxidative phosphorylation. Polymorphonuclear leukocytes also generate free radicals as part of the inflammatory response. Although transient, species such as the superoxide radical (O₂·⁻) and the hydroxyl radical (OH·) can overwhelm natural antioxidant defenses, altering proteins, nucleic acids, and lipids (lipid peroxidation). This can result in cell injury or death, subsequent tissue damage, and ultimately, a chronic disease state.¹

Free radical–mediated damage has been characterized in the pathogenesis of more than 100 disorders including stroke, atherosclerosis, myocardial infarction, autoimmune diseases, and nervous system disorders.¹ Studies in our laboratory have implicated free radicals in otitis media (OM) in animal models^2-5 and humans.⁶ Neutrophils, in their respiratory burst, and Streptococcus pneumoniae, the most common pathogen in acute OM,⁷ are thought to generate the damaging species in this infection.

The previous findings that free radicals are indeed involved in OM prompt a search for them in other related diseases. Although clinically separate, rhinosinusitis and OM share a common pathogenesis. Both are closed-space infections involving the blockage of a natural orifice (the eustachian tube in OM and the sinus ostia in rhinosinusitis) leading to stasis and subsequent bacterial infection.

No recent studies in the English-language literature have assessed for free radical–mediated damage in human rhinosinusitis. Tissue samples from patients undergoing functional endoscopic sinus surgery to treat chronic rhinosinusitis were collected and the lipid peroxide (LPO) content of control and diseased mucosa was determined.

A series of 13 consecutive patients recruited over a 10-week period underwent functional endoscopic sinus surgery to treat chronic rhinosinusitis. All patients signed a standard informed consent form allowing for research evaluation of surgical tissues. None of the patients were taking systemic steroids prior to surgery. Samples of the diseased mucosa (chosen for obvious signs of inflammatory thickening and erythema) and corresponding control samples from healthy-appearing mucosa in the nasal septum, inferior turbinate, or ethmoid sinus were obtained for each case. Because the study dealt with removed tissue and did not affect the patients themselves, the institutional review board granted exemption from the formal approval process.

The tissues were placed in a solution containing cold 0.1M triethanolamine-buffered saline with 1mM phenylmethylsulfonyl fluoride, 1mM EDTA, and 1mM dithioerythritol. Samples were mechanically homogenized with a glass grinder, disrupted with a sonicator for 3 seconds, and centrifuged at 1000g for 5 minutes at 4°C. The supernatant was frozen in a light-deprived environment until all samples were collected and ready to be analyzed. Two separate aliquots from each of the control and diseased supernatants were processed. Sample protein concentration was determined colorimetrically with the BCA protein assay (Pierce Endogen, Rockford, Ill) as described by Smith et al.⁸

Because molecules such as the superoxide and hydroxyl radicals are transient species that are difficult to measure directly, the amount of LPO, a by-product of free radical damage to cell membranes, was quantified. The Determiner LPO-CC kit (Kamiya Biomedical, Thousand Oaks, Calif) uses the hemoglobin-catalyzed stoichiometric reaction of hydroperoxides with 10-N-methylcarbamoyl-3,7-dimethylamino-10-H-phenotholthiazine (MCDP) to form methylene blue, which can be measured with a spectrophotometer.

In a light-deprived environment, samples were mixed with kit reagent 1 (ascorbic oxidase and lipoprotein lipase) and incubated at 37°C for 10 minutes. Kit reagent 2 (MCDP) was then added and the samples were reincubated at 37°C for 15 minutes. Absorbance was measured at 675 nm using a microplate reader, and the LPO content (expressed in nanomoles per milliliter) was determined relative to a known cumene standard provided with the kit. Each of the 2 aliquots from the control and diseased groups was processed in triplicate. Therefore, a total of 12 LPO concentrations were obtained (6 for the control and 6 for the diseased tissue) for each patient. Mean control and diseased values were then divided by their respective sample protein concentrations to yield LPO/protein ratios for each patient. A paired t test with 12 df was used to ascertain whether a statistically significant difference in free radical–mediated damage between normal and chronically inflamed tissue was present.

Figure 1 depicts the LPO/protein ratios from control and diseased tissue samples for each of the 13 patients. There was not a statistically significant difference in the degree of free radical–mediated damage between healthy and inflamed tissue (95% confidence interval, −3.00 × 10^-5 to 2.94 × 10^-5). The mean LPO/protein ratio for healthy tissue was 3.52 × 10^-5 with an SE of 1.57 × 10^-7, while that for the diseased tissue was 3.49 × 10^-5 with an SE of 1.73 × 10^-5 (Figure 2).

Biopsy specimens of the nasal turbinates were taken as controls in most patients (10/13) while the ethmoid sinus was the most frequent biopsy site for diseased tissue (10/13). One patient (No. 3) had diseased and control biopsy specimens taken from inflamed and healthy areas (as determined by the operating surgeon), respectively, of the ethmoid sinus.

The data obtained are the result of a pilot study of free radical–mediated damage in human patients with chronic rhinosinusitis. There seems to be no difference between healthy and inflamed tissue. This may be due to the chronicity of disease in the study patients. Previous work in our laboratory has demonstrated that guinea pigs with experimentally induced OM have statistically significant decreases in lipoperoxidation at 30 days after infection vs 5 days after infection⁴ and that levels of hydrogen peroxide, a by-product of oxidative metabolism and significant mediator of free radical–induced damage, are highest 24 hours after infection.³ All 13 patients in the present study had chronic rhinosinusitis, which, by definition, necessitates a symptom duration of 12 weeks or longer.⁹ Perhaps free radical–mediated damage in rhinosinusitis wanes with time because antioxidant defense mechanisms, which are overwhelmed in the acute stage of disease, catch up as inflammation persists and time allows for tissue repair. Moreover, the nature of the inflammatory response is known to be different in acute vs chronic disease. Acute sinusitis leads to a substantial influx of neutrophils,¹⁰ which generate free radicals during their respiratory burst; however, studies evaluating the histopathologic characteristics of chronic sinusitis have shown a predominantly lymphocytic response, with only occasional neutrophils present.^11,12 Thus, lipoperoxidation may have occurred initially, during the acute phase of rhinosinusitis, but not during the chronic inflammation that the study subjects were known to endure. As mucosal tissue healed from the initial insult, perhaps free radical–induced damage decreased as well.

Two possible experimental design aspects may also account for the lack of difference in lipoperoxidation. First, the use of a control specimen from the same patients with diseased tissue eliminated the possibility of confounding but also relied on the surgeon's ability to choose healthy tissue based on visual inspection. Given the chronicity of the patients' disease, it is certainly possible that tissue that did not appear inflamed at the time of surgery was affected prior to the procedure and may have suffered free–radical mediated damage that was subsequently detected by the LPO assay. This raises the possibility that control samples were not representative of healthy tissue.

Second, tissue samples were initially processed and frozen until further analysis was possible. The freezing process can generate radicals capable of damaging biological molecules and, in fact, this is a common problem in the production of proteins for biological use.¹ The prolonged freezing of the samples had an uncertain effect on final results. It is important to note, however, that freezing was also used in the storage of samples during previous experiments in our laboratory, experiments that did demonstrate a difference in free radical–mediated damage.^2-6

Although this study showed no statistically significant difference in LPO content between healthy and diseased tissue in chronic rhinosinusitis, perhaps the difference, if it exists at all, is smaller and therefore more difficult to detect than that which has been shown in OM. While the neutrophil respiratory burst is a part of the inflammatory response and contributes to free radical formation in OM and rhinosinusitis, S pneumoniae, the other source of free radicals in OM, may not play as large a role in chronic rhinosinusitis. Though it is one of the most common pathogens in acute sinusitis, pneumococcus is relatively rare in cultures from patients with chronic sinusitis. A recent study of 174 patients with chronic maxillary sinusitis grew S pneumoniae in only 0.5% of cultures, and a review of the literature failed to implicate it as a predominant organism in 7 of the 9 most recently published articles on the subject.¹³

In conclusion, the present data represent a preliminary investigation into the role of free radicals in chronic rhinosinusitis. Further studies involving larger numbers of samples and the use of an animal model for rhinosinusitis are planned.

Accepted for publication February 27, 2002.

Corresponding author: Joseph Haddad, Jr, MD, Columbia-Presbyterian Medical Center, Babies' and Children's Hospital, 3959 Broadway, Room 501N, New York, NY 10032 (e-mail: jh56@columbia.edu).