TEXT

Top Abstract Text References

Acute respiratory illnesses contribute significantly to acute morbidity, physician visits, and absenteeism from work and school (15). The common cold is a viral rhinosinusitis (VRS), with a computerized tomography (CT) scan showing sinus involvement in 87% of adults with colds (5). Acute community-acquired bacterial sinusitis (ACABS) complicates a small proportion of VRS, with reports ranging from 0.5 to 2% (1, 2). The "gold standard" for diagnosis of ACABS is sinus aspirate culture yielding an identifiable bacterial strain. However, sinus aspirate is not appropriate for routine clinical use, and a readily available, simple test for ACABS that may require antibiotic treatment is badly needed.

In addition, allergy and eosinophil involvement in some patients with chronic sinusitis is well recognized (7, 8, 12, 13). The neutrophil marker lactoferrin (LF) and the eosinophil marker eosinophilic cationic protein (ECP), respectively, are potentially useful markers of neutrophilic inflammation in fecal, cervicovaginal, and sputum samples (3, 11, 14, 16, 17) and of eosinophilia in nasal washings (9).

We therefore compared LF titers and ECP levels in nasal secretions or washes from healthy adults, volunteers with experimental rhinovirus colds, adults with natural colds, and adults with presumed ACABS.

We examined nasal secretion specimens (blown into a plastic wrap and weighed for dilution) from 9 healthy adults, 9 adults with natural colds with a history of <7 days of symptoms, and 16 adults who presented to one of two primary-care outpatient clinics with a history of at least 7 days of stable or worsening respiratory symptoms (suggested as a clinical indicator of ACABS) (4). We also examined nasal washings from 32 previously healthy volunteers exposed to an experimental rhinovirus (Hank's strain), a subset of whom went on to acquire colds (n = 19), while a second subset did not (n = 13). No participants had a history of chronic sinusitis, active or chronic allergies, or any other chronic upper respiratory disease. The study was approved by the human investigation committee at the University of Virginia, and informed written consent was obtained.

Nasal washings were collected immediately before rhinovirus inoculation and at day 4 (average day of peak symptoms) of experimental rhinovirus colds as described previously (6). Nasal secretion specimens were diluted 50-fold, in Leukotest diluent (0.1% sodium azide buffer; TechLab Inc., Blacksburg, Va.) with 0.1% Triton 100 detergent to lyse the neutrophils and eosinophils and release their granular contents, and were assayed with the Leukotest LF latex agglutination assay, with positive specimens further diluted 2-fold to determine a final titer. ECP was measured using a monoclonal antibody-based fluorometric assay (Pharmacia CAP system; Pharmacia Diagnostics, Uppsala, Sweden).

Because nasal wash specimens (from volunteers before and after experimental rhinovirus colds) were collected with an unspecified amount of saline wash, and because plasma urea freely diffuses to equivalent mucosal levels (10), urea concentrations in specimens and sera were determined by using Sigma Diagnostics kit no. 66-UV, a spectrophotometric assay based on the enzymatic hydrolysis of urea, and were used to calculate the nasal washing dilution for LF and ECP levels.

Fisher's exact test was used to compare proportions. Student's t test was used to compare ordinal data. All P values are two-sided.

Subjects with symptomatic experimental rhinovirus colds had a slight increase in the mean nasal wash LF titer of 1:82 (standard error [SE], ±55) compared to a baseline before inoculation of 1:25 (±14) (P < 0.04). Subjects who were not infected had LF titers of 1:69 (±46) and 1:41 (±18) at days 0 and 4, respectively (P > 0.05). In the infected and ill group, 12 of 19 (63%) had a mild increase (i.e., none to 1:400) in the nasal wash LF titer after inoculation with rhinovirus, while only 4 of 13 (31%) subjects who were not infected had any increase in the nasal wash LF titer. Nasal secretion LF titers were almost identical when asymptomatic adults were compared with subjects with natural colds. Eight (89%) of nine healthy adults and seven (78%) of nine subjects with symptomatic natural colds had nasal secretion LF titers of 1:200, and none had nasal secretion LF titers of >1:400. In contrast, as summarized in Fig. 1, 11 (69%) of 16 subjects with presumed ACABS had nasal secretion LF titers of 1:400, compared with only 3 of 18 (17%) controls and subjects with natural colds (P < 0.01 by Fisher's exact test).

View larger version (21K): [in this window] [in a new window]   FIG. 1.   Reciprocal LF titers for controls and various respiratory infections. Reciprocal LF titers in nasal secretions or sputum specimens are most often 400 in subjects with presumed ACABS, bronchitis, or pneumonia (19 of 24; 79%), while they are most often <400 in healthy subjects and those with both natural and experimental rhinovirus colds (77 of 80; 96%) (P < 0.001 by Fisher's exact test). Eleven control saliva and 8 purulent sputum specimens are included from reference 11 for comparison.

ECP concentrations for all specimens available for study are shown in Fig. 2. Levels of ECP of >200 ng/ml were considered elevated. While none of the 12 subjects with experimental rhinovirus colds had a baseline nasal secretion ECP concentration of >200 ng/ml before infection, all 22 available nasal secretion specimens from subjects with natural colds or presumed ACABS, and specimens from 5 (42%) of 12 subjects with experimental rhinovirus colds, had ECP levels of >200 ng/ml. Altogether, 27 (79%) of 34 subjects with presumed ACABS or colds (natural or experimental) had ECP levels of >200 ng/ml (versus 0 of 12 baseline controls; P < 0.001). Furthermore, 10 (62%) of 16 specimens from subjects with presumed ACABS showed strikingly elevated ECP levels (>3,000 ng/ml, including 3 specimens with LF titers of <400), while only 3 of 18 subjects with colds (natural or experimental) and none of the 12 baseline controls had ECP levels of >3,000 ng/ml (P < 0.01).

View larger version (17K): [in this window] [in a new window]   FIG. 2.   ECP levels in nanograms per milliliter for all specimens available for study, obtained by using either urea (for nasal washes) or specimen weight (for nasal secretions) as noted in the text. Twenty-seven (79%) of 34 subjects with presumed ACABS or colds (natural or experimental) had ECP levels of >200 ng/ml, while none of the 12 baseline controls did (P < 0.001 by Fisher's exact test). Furthermore, 10 (62%) of 16 specimens from subjects with presumed ACABS had strikingly elevated ECP levels (>3,000 ng/ml), while no specimens from 18 subjects with colds (natural or experimental) and none of the 12 baseline control specimens did (P < 0.001 by Fisher's exact test).

We conclude that LF measurement in patients with acute respiratory illness may prove to be useful in distinguishing uncomplicated colds from colds complicated by ACABS. Furthermore, the involvement of eosinophils, as indicated by ECP concentrations in nasal secretions, suggests that allergy may also be involved in a substantial portion of colds and sinusitis syndromes. These pilot studies should be extended to patients with sinus puncture and culture as well as antibiotic responses to define the ultimate place for nasal LF and ECP measurement in patient management decisions, such as when antibiotics may be indicated.