The pathogenesis of the disease is not well known. Several anecdotal reports have noted beneficial results of sulfamethoxazole-trimethoprim in the treatment of refractory Wegener granulomatosis or limited Wegener granulomatosis localized to the respiratory tract [7-10]. These findings have raised speculation regarding a possible role for infectious agents in the induction of disease activity in Wegener granulomatosis. Recently, it was noted that patients with Wegener granulomatosis frequently have symptoms of a respiratory tract infection at the time of ANCA titer elevation [4]. Previous studies also found that many patients had respiratory tract infections either at the onset of the disease or before a relapse [11, 12]. Virtually every patient with Wegener granulomatosis has secondary infections of the paranasal tissues, predominantly with Staphylococcus aureus, and infections invariably respond to antistaphylococcal antibiotics [2]. Nasal carriage of S. aureus is considered a risk factor for S. aureus infections [13-16]. No prospective studies have been done on the frequency of these infections or the relation of nasal carriage rate of S. aureus to disease activity or levels of ANCA in patients with Wegener granulomatosis. We examined several possible risk factors for relapse of Wegener granulomatosis.

Methods

Top Methods Results Discussion Author & Article Info References

All consecutive patients with biopsy-proven Wegener granulomatosis visiting our outpatient clinic from January 1988 to July 1991 entered the study. The diagnosis of Wegener granulomatosis was based on clinical and histologic criteria [2]. At each visit, usually every 4 to 6 weeks, patients were evaluated for signs and symptoms attributable to active Wegener granulomatosis and infections; blood samples were obtained for ANCA determinations; a swab culture of the anterior nares was taken; and medication was noted, with special emphasis on current or recent use of antibiotics. Because antibiotics can influence the results of nasal cultures, results of cultures taken during or within 4 weeks after a course of antibiotics were disregarded, as were cultures taken during hospitalization because the carrier rate may be enhanced by hospital stay [13]. No attempts were made to eliminate S. aureus carriage, although proven or suspected infections were treated with appropriate courses of antistaphylococcal antibiotics.

Relapses of Wegener granulomatosis were treated, and immunosuppressive agents were tapered according to a protocol described previously [3]. Briefly, patients were initially treated for active Wegener granulomatosis with cyclophosphamide, 2 mg/kg body weight daily, and prednisolone, 1 mg/kg daily. Doses of cyclophosphamide were adjusted to maintain the leukocyte count above 4.0 x 10⁹/L. In the absence of signs of active disease, the daily dose of cyclophosphamide was tapered every 3 months by 25 mg. Daily prednisolone was tapered after 6 weeks by 10 mg every 2 weeks until the dose reached 30 mg and thereafter by 5 mg every 2 weeks.

Given the possible beneficial influence of cotrimoxazole in Wegener granulomatosis, patients receiving prolonged continuous treatment with antibiotics, arbitrarily defined as more than 6 weeks, were excluded. To establish true stable nasal carriage of S. aureus, only patients with at least 1 year of follow-up were included in the analysis.

Patients were considered to have relapsed when they satisfied one of the following criteria in which the manifestations in question had to be new or recurrent findings [4]:

1. Progressive glomerulonephritis, a decrease in renal function of 30% or more within 3 months in combination with (microscopic) hematuria or evidence of acute necrotizing lesions on renal biopsy.

2. Pulmonary infiltrates with or without cavitation with rising C-reactive protein levels in the absence of infection or signs of malignancy.

3. Sinusitis, otitis, ulceration of nasal mucosa, or proliferative mass, in combination with necrotizing granulomatous inflammation on biopsy.

4. Miscellaneous (progressive mononeuritis multiplex, cranial nerve palsy, cerebral vasculitis, necrotizing scleritis, orbital pseudotumor, progressive tracheal stenosis with active disease on biopsy, peripheral gangrene, necrotizing vasculitis of medium-sized or small muscular arteries).

An upper-airway infection was considered present when patients had clinical signs involving the pharynx, trachea, nasal and paranasal tissues, and ear, in combination with a positive culture or serologic evidence, and when patients had clinical signs suggesting infection, without microbiologic evidence but with a prompt response to appropriate antibiotic treatment. Viral serologic tests were done only when clinically indicated, as were serologic tests for Mycoplasma, Chlamydia, and Legionella species. Urinary tract infections were defined by leukocyturia in combination with bacterial concentrations of 10⁵ colony-forming units per mL or greater.

Patients were considered to be chronic nasal carriers when 75% or more of their nasal cultures grew S. aureus [14] and nonchronic carriers when less than 75% (or no) nasal cultures grew S. aureus. Patients were also grouped according to their ANCA status during the study. Patients were considered persistently positive for ANCA when all ANCA measurements were positive during the study and intermittently positive when periods with positive ANCA test results were interrupted by at least two consecutive negative readings. Patients were considered persistently negative for ANCA when all consecutive ANCA measurements were negative from 6 months after the previous period with disease activity [6]. Destructive nasal lesions were defined as visible perforations of the nasal septum or saddle-nose deformity.

Nasal cultures were taken by firmly rotating a sterile cotton-tipped swab (Transwab; Amies Med Elpo, Italy) in each anterior naris. The swabs were inoculated on 5% sheep-blood agar and salt mannitol agar (Oxoid; Basingstoke, England) for 48 hours at 35 °C. Isolates were identified as S. aureus by the typical appearance of the colonies that were coagulase positive by slide coagulase testing and the ability to cleave DNA. Antineutrophil cytoplasmic antibodies were measured by indirect immunofluorescence on ethanol-fixed granulocytes [17]. Antineutrophil cytoplasmic antibodies were designated as c-ANCA when the fluorescence pattern was granular with a decrease in fluorescence intensity toward the periphery of the cells and as p-ANCA when a perinuclear-to-nuclear pattern was observed. The specificity and sensitivity of c-ANCA determinations in our laboratory for active Wegener granulomatosis have been previously reported [3]. In addition, the antigenic specificity of ANCA for either proteinase-3, myeloperoxidase, and elastase was tested by ELISA as previously described [18].

Statistical Analysis

Differences in frequencies of categorical variables between groups were studied by the chi-square test. The continuous variables were tested with the Wilcoxon rank-sum test. Disease-free interval curves were calculated using the Kaplan-Meier method [19]. The first relapse during the study period was counted from the most recent period of disease activity, either diagnosis or a prestudy relapse. Patients without relapse were censored at the end of the study. Differences between groups in disease-free interval were studied by the log-rank test [20]. Confounding influences were tested by tests for trend and, when statistically significant, the log-rank test was done after stratification for the confounding variable [20]. Stepwise Cox proportional-hazards regression analysis (BMDP statistical software, BMDP, Inc., Cork, Ireland) with disease-free interval as the dependent variable was used to adjust the observed effect of S. aureus carrier status for other potential prognostic factors and to determine the independent influence of these other factors on disease-free interval [21]. Other variables included age, sex, disease-free interval before the start of the study, time since diagnosis, renal function measured as creatinine clearance (used as both a continuous and a categorical variable), nasal structure, number of documented infections during the study, history of previous relapses, and the time at which the patient was diagnosed as having Wegener granulomatosis (before or during the study). The termination of immunosuppressive therapy after the previous period of disease activity was included as a time-dependent covariate. Hazard ratios are reported as relative risks with 95% CIs computed from the exponential in the regression model using only covariates whose coefficients had a P value 0.10. Association of categorical variables was assessed by chi-square analysis, and associations were expressed by relative risk and its 95% CI [22]. A two-tailed P value < 0.05 was considered statistically significant.

Results

Top Methods Results Discussion Author & Article Info References

Of 71 patients who entered the study, 57 were analyzed, including 33 with an established diagnosis of Wegener granulomatosis before January 1988. The excluded patients included 11 patients initially referred for diagnosis or treatment but who requested follow-up at the referral hospital, 1 patient who died from gram-negative septicemia and pneumonia 8 months after being diagnosed with Wegener granulomatosis, and 2 patients who received long-term antibiotic treatment for recurrent urinary tract infection within a year of study entry. The excluded patients did not differ statistically from the patients analyzed with respect to clinical characteristics.

A mean of 22.6 nasal cultures per patient (range, 10 to 31 cultures) were taken. Forty-two patients (74%; 95% CI, 60% to 84%) had one or more nasal cultures that grew S. aureus, and 36 (63%; CI, 49% to 76%) were chronic nasal carriers. The six patients with positive cultures who were not chronic nasal carriers had a low frequency of positive cultures (range, 4.5% to 15.0%). The mean rate of positive nasal cultures in chronic nasal carriers was 96.5% (range, 83.3% to 100%). In 23 of the 36 chronic nasal carriers (64%; CI, 46% to 79%), all nasal cultures were positive for S. aureus. The number of cultures obtained from chronic nasal carriers and nonchronic carriers of S. aureus did not differ statistically (22.4 compared with 22.9, respectively; P > 0.2). Based on only the first eight nasal cultures taken per patient, all but 2 of the 36 chronic nasal carriers of S. aureus would have been classified correctly and none of the nonchronic carriers would have been misclassified.

No statistical differences in clinical characteristics were found between the groups with and without chronic nasal carriage of S. aureus Table 1 and Table 2. Chronic nasal-carrier status for S. aureus was not associated with lower renal function (relative risk, 1.53; CI, 0.54 to 2.78), presence of previous relapses (relative risk, 1.08; CI, 0.70 to 1.65), occurrence of upper airway infection during the study (relative risk, 1.20; CI, 0.75 to 1.92), abnormal nasal structure (relative risk, 1.08; CI, 0.70 to 1.65), or use of immunosuppressive medication at the start of the study (relative risk, 0.84; CI, 0.55 to 1.29). Long-term hemodialysis treatment and insulin-dependent diabetes mellitus, well-known risk factors for carriage of S. aureus, were present only in three patients and one patient, respectively. Chronic nasal carriers of S. aureus did not differ statistically from nonchronic carriers with respect to the number of previous relapses, duration of disease and latest disease-free interval, and use of immunosuppressive drugs in the 33 patients with Wegener granulomatosis at study entry, nor was chronic nasal carriage of S. aureus more frequent in this group (relative risk, 1.02; CI, 0.68 to 1.52).

Twenty-three of the 57 patients developed 32 relapses of Wegener granulomatosis during the study. Fifteen patients had one relapse, 7 had two relapses, and 1 patient had three relapses. Of the 36 patients with chronic nasal carriage of S. aureus, 21 (58%; CI, 41% to 74%) had one or more relapses, whereas only 2 of 21 patients (10%; CI, 1% to 30%) without chronic nasal carriage had a relapse (P < 0.001, Figure 1. Nasal upper-airway lesions, progressive glomerulonephritis, and pulmonary lesions were most frequently associated with relapse (Table 3). Renal function, arbitrarily classified as a creatinine clearance above or below 60 mL x min^–1, was an independent risk factor [P < 0.02]. No such influence was observed for the other variables. In the group of 23 patients with relapses, 9 had a proven or clinically suspected period of upper-airway infection during the 6 months preceding the relapse. Twelve patients had a relapse some time after withdrawal of immunosuppressive medication (median, 11 months; range, 2 to 96 months), and 11 patients had a relapse during the period that the dose of immunosuppressive agents was being tapered according to the treatment protocol.

View larger version (13K): [in this window] [in a new window]   Figure 1. Disease-free interval and carrier status. Disease-free interval of 57 patients with Wegener granulomatosis grouped according to Staphylococcus aureus carrier status. The time of the disease-free interval was counted from the beginning of the most recent period of disease activity (either initial diagnosis or relapse; P < 0.001).

Fifty-four of the 57 patients (95%) had c-ANCA with specificity for proteinase-3 at some time. Of the remaining 3 patients, 2 had p-ANCA with specificity for myeloperoxidase, and 1 patient never had demonstrable ANCA. The median number of ANCA determinations per patient during the study was 29 (range, 11 to 49). All but one relapse occurred in patients with persistently or intermittently positive c-ANCA determinations (Figure 2). Chronic nasal-carrier status for S. au-reus was associated with persistent or intermittent c-ANCA positivity (relative risk, 1.84; CI, 1.04 to 3.26). Renal function with a creatinine clearance below 60 mL x min^–1 was associated with persistent absence of c-ANCA (relative risk, 1.98; CI, 1.15 to 3.41).

View larger version (12K): [in this window] [in a new window]   Figure 2. Disease-free interval and c-ANCA status. Disease-free interval according to the course of c-ANCA during the study (n = 54). The time of the disease-free interval was counted from the beginning of the most recent period of disease activity (either initial diagnosis or relapse). (c-ANCA-negative versus intermittently or persistently c-ANCA-positive, P < 0.001). Dashed line = c-ANCA-negative (n = 21); Broad line = intermittently c-ANCA-positive (n = 21); thin solid line = persistently c-ANCA-positive (n = 12).

Analysis of the disease-free interval with the proportional-hazards model identified chronic nasal carriage of S. aureus (P = 0.009), renal function classified by creatinine clearance above or below 60 mL x min^–1 (P = 0.01), and a history of previous relapses of Wegener granulomatosis (P = 0.06) as covariates whose coefficients had a P value 0.1. The relative risks for the occurrence of relapse of Wegener granulomatosis in the proportional-hazards regression model including these three variables were 7.16 (CI, 1.63 to 31.50) for chronic nasal carriage of S. aureus, 2.94 (CI, 1.27 to 6.67) for a creatinine clearance above 60 mL x min^–1, and 1.33 (CI, 0.98 to 1.78) for a history of previous relapses. The inclusion of proven and clinically suspected upper-airway infections, either as a categorical (absent or present) variable or as infection rate (number of infections per year), in the proportional-hazards regression model did not lead to a significant contribution (P > 0.2 for both), nor did upper-airway infections contribute significantly in any combination with one or more covariates.

Discussion

Top Methods Results Discussion Author & Article Info References

An important role for an infectious agent or agents in the induction of Wegener granulomatosis was first suggested by Friedrich Wegener in 1936 in the original descriptions of the disease [29]. He proposed that the disease is triggered by an infection of the upper respiratory tract and that all the ensuing sequelae are an allergic reaction to this infection. Others also have suggested a role for infection in triggering relapses of Wegener granulomatosis [12]. In our study, chronic nasal carriage of S. aureus and not the occurrence of clinically apparent infections was associated with relapses of the disease. It is possible, however, that self-limiting viral or bacterial infections were missed in our study setting because they may not have been reported consistently by the patients. Also, serial viral serologic evaluation was done only when clinical signs of infection were observed or reported.

The reported frequency of chronic or stable nasal carriage of S. aureus in a healthy population varies from 20% to 35%, depending on the criteria used and the population tested [13, 23-27]. Minor alterations in nasal structure may influence the propensity to carriage of S. aureus [28], so nasal carriage of S. aureus in patients with Wegener granulomatosis is possibly facilitated by underlying nasal damage or mucosal alterations resulting from the disease. However, visible changes in nasal structure resulting from Wegener granulomatosis were equally found in nonchronic carriers and chronic carriers. Known risk factors for chronic nasal carriage of S. aureus such as diabetes mellitus or chronic hemodialysis treatment were only present in a few patients.

We also looked for other factors that might predict the relapse rate of Wegener granulomatosis. In the group of patients with Wegener granulomatosis and a positive c-ANCA test at any time during the course of their disease (comprising 95% of the patients we studied), relapses were nearly totally confined to the group with persistent or intermittent presence of detectable c-ANCA in serial determinations. However, a substantial proportion of the patients with persistent or intermittent c-ANCA determinations did not have a relapse. The temporal relation between changes in c-ANCA and disease activity, which has been reported in several studies [3-6, 17], but not in others [30], might be an epiphenomenon or may indicate a pathophysiologic relation between ANCA and disease activity. The association of chronic nasal carriage of S. aureus with intermittently or persistently positive c-ANCA test results and with relapse of Wegener granulomatosis suggests that S. aureus may have a role in the induction of disease activity, although it is possible that low-grade or subclinical disease activity of Wegener granulomatosis by inducing necrosis in the nasal tissues facilitates carriage of S. aureus. Theoretically, low-grade infection with S. aureus might prime neutrophils, resulting in the expression of proteinase-3 and myeloperoxidase, the target antigens of ANCA, on their cell surface. In the presence of ANCA, primed neutrophils may be further activated to release lysosomal enzymes and produce oxygen radicals leading to a more clinically apparent chronic inflammatory process [31, 32]. In six patients with re-activation of Wegener granulomatosis, bronchoalveolar lavage studies have shown neutrophilic alveolitis in the absence of infectious agents but in the presence of c-ANCA [33].

Staphylococcus aureus possesses some features that suggest a more direct causal role in facilitating autoimmune phenomena. Several toxins of S. aureus have been shown to stimulate B cells and probably T cells in an unrestricted and T-helper cell-independent way [34-37]. Cell-wall components of S. aureus, such as peptidoglycan, are effective B-cell mitogens and polyclonal activators, resulting in, among others, the activation of autoreactive B cells that may produce ANCA [38-40]. Direct stimulation by S. aureus of neutrophils in vitro has also been documented [41]. In addition, S. aureus produces proteinases with activity against human proteinase inhibitors like ₁-antiprotease [42]. ₁-Antiprotease has been shown to be the main inhibitor of proteinase-3, and low levels may facilitate the interaction between proteinase-3 and proteinase 3-reactive lymphocytes as well as the proteolytic activity of proteinase-3 resulting in persistent inflammatory activity [43, 44]. This, again, might further trigger the occurrence of relapses. Whether these effects shown in vitro occur locally in the upper airways and contribute to the development of disease activity in Wegener granulomatosis by either S. aureus colonization or infection remains speculative.

Decreased renal function was associated with fewer relapses of Wegener granulomatosis, independent of the nasal carrier status of S. aureus. Previous small series and case reports have produced conflicting findings [45-47], and prospective data regarding this issue are lacking.

In conclusion, chronic nasal carriage of S. aureus in a group of 57 patients with histologically proven Wegener granulomatosis followed for 1 to 3.5 years identified a group more prone to relapse. The causal relation and mechanisms of this association remain speculative. Studies aimed at eliminating S. aureus carriage are warranted to see if relapses may be prevented.

Part of this study was presented at the 4th International Workshop on ANCA and 2nd Colloquium on Wegener granulomatosis held 29 and 30 May 1992 in Lubeck, Germany.