Primary Sjögren's syndrome (pSS) is a systemic disease characterized by a chronic autoimmune inflammatory process. The most relevant anatomopathologic finding that characterizes it is the presence of a focal lymphocytic infiltration of the exocrine glands, being one of the main diagnostic criteria. The inflammation leads to glandular destruction and atrophy that evolves towards mucosal dryness, mainly ocular (keratoconjunctivitis sicca), oral (xerostomy), tracheal (xerotrachea) and vaginal (vaginal xerosis).1,2 The lungs share with the exocrine glands a similar structure; therefore, their affectation can become evident during the course of the syndrome. The most frequent respiratory symptoms are chronic cough and dyspnea,3,4 which can reflect an affectation of the bronchial tree or of the pulmonary parenchyma.5–7 In the literature, the series of cases that have studied the physiopathology of the cough and dyspnea in pSS implicate an inflammatory pattern of the pulmonary interstitium and/or a concomitant affectation of the peripheral airway, mainly,3,8,9 although a poor correlation has been reported among symptoms, histopathology, lung function and radiology in pSS.9–11 On the other hand, it is unknown whether the inflammation present in the bronchial lumen, which has not been studied fully, can reflect either an early or a late stage in the natural history of the syndrome in those patients who meet the 2002 consensus diagnostic criteria12 and, therefore, if histopathology should instead be considered the gold standard pattern of the disease.2 The objectives of this study are to describe the clinical characteristics of the lung function and the inflammation present in the bronchial lumen of patients diagnosed with pSS, and to analyze whether the inflammatory profile, obtained in a non-invasive manner by means of induced sputum analysis, is associated with the presence of respiratory symptoms.

Thirty-six patients were included in the study, with a mean age of 63, 10.4 (SD), 92% of whom were women. 89% presented labial biopsy with lymphocytic infiltration, 31% Ro/La autoantibodies and 50% FR/ANA autoantibodies. The mean time since the onset of the first pSS symptom was 10, 5 (SD) years and since diagnosis 6, 5 (SD) years. Table 1 shows the distribution of frequencies of the most relevant symptoms in the presentation of the syndrome, the most predominant symptoms and those most frequently combined with respiratory symptoms (clearing of the throat, dyspnea or cough). Clearing of the throat was present in 58% of the patients at some time during their evolution, cough and mild dyspnea (MRC scale=2) in 42% of the patients. The mean time with cough was 24, 40 (SD) months.

In 10 (28%) of the patients, some kind of radiological alteration of the thorax was detected with either local or diffuse affectation (interstitial or attenuation of the radiological lung density), and an associated alteration was confirmed in the volumes and/or in the diffusion in half (n=5). In 7, the sputum presented lymphocytosis higher than 2.6%.

The lung function tests were carried out and the inflammatory profile found in the bronchial lumen, both in the entire sample as well as in the two groups of patients considered, depending on whether the sputum was considered normal or pathological, are shown in Table 2. There were no relevant alterations in the lung function in the entire study sample, except for the residual volume that was higher in the group of subjects with pathological sputum.

Of all the sputum analyzed (n=35), 69% (n=24) were considered pathological, while the remaining 31% (n=11) were normal. The patients who presented pathological inflammatory profiles did so exclusively at the expenses of the lymphocytes (>2.6%). There were no statistical differences between the two groups regarding time elapsed since the appearance of the first symptom: 11, 6 (SD) in those who presented pathological sputum compared with 9, 3 (SD) years in those with normal sputum. The same was true for the time elapsed since diagnosis: 7, 5 (SD) in those who presented pathological sputum compared with 6, 4 (SD) years in patients with normal sputum.

Table 3 shows the characteristics of the cough and the bronchial response with bronchodilation or methacholine, according to the inflammatory profile found in the bronchial lumen.

In all the sputum samples except three, the lipophage count was analyzed (macrophages with presence of lipid inclusions). 42% (n=14) presented a pathological index higher than 15 (range of count 0 to 400), distributed in 55% (n=12) of the pathological sputum (n=22) compared with 18% (n=2) of the normal sputum (n=11), P=.05.

The majority of patients with pSS present respiratory symptoms (clearing of the throat, cough and dyspnea) at some time during the evolution of their disease, along with other concomitant symptoms of extraglandular affectation (ocular, oral, etc.). These are more predominant than the respiratory symptoms in the evolutionary course of the disease, especially at the beginning. The prevalence and mean time from the onset of the systemic and respiratory symptoms described in this series coincide with those reported in the literature, with dyspnea and cough being the most relevant among the respiratory symptoms.1,3,5,18 In our series, chronic cough was the respiratory symptom that presented a more prolonged duration over time and was combined with other symptoms of ocular or oral dryness. It was not the initial or predominant symptom among those detailed, although its verification should lead to the suspicion of pSS diagnosis in the management of chronic cough.19 Cough has been postulated as a main symptom of tracheal affectation above other respiratory symptoms in the glandular affectation of pSS. It may be accompanied by dyspnea in the presence of radiological affectation or alteration of the lung function in the extraglandular form, resulting in parenchymatous and bronchiolar affectation.9 Mention should be made that, in the patients we analyzed, in those who coughed more and in those whose cough lasted longer, sputum was normal (Table 3).

On the other hand, in the entire series studied the radiological and lung function affectation was mild in some patients, although its detection implicates a close follow-up of the symptoms. These findings agree with the descriptions of the literature and with the limited radiological and lung function associations over time reported for pSS.8,10,11,20 It should also be remarked that in the majority of this subgroup of patients with radiological or pulmonary affectation, the sputum presented lymphocytosis.

33% of the cases presented positive bronchial challenge with methacholine, and 17% presented positive bronchodilator test. Together, these percentages should be considered as high overall, and possibly reflect a tracheobronchial immunological affectation with hyperresponse, as has already been suggested by the series of La Corte et al.21 and Gudbjörnsson et al.22 Although when analyzed separately these hyperresponse tests do not demonstrate a different profile in the sputum, when the presence of one or the other is confirmed and they are grouped for analysis, the predominant profile in the sputum is lymphocytosis (Table 3). Therefore, this hyperresponse would be that of a non-specific response of the airway dysfunction with a difficult physiopathological translation, but it would implicate the affectation of the glands of the tracheobronchial mucosa,23 the submucosa or the alveolar bed.3,7

Among all the results of our study, what is especially relevant is the inflammatory profile found in the non-invasive samples collected from the induced sputum, which were two-thirds pathological. The presence of lymphocytosis in the majority of the subjects of our series is in accordance with the alveolar inflammatory cell affectation predominant at the parenchymatous and bronchioloalveolar level published for this syndrome.3–5,7 Thus, we can affirm that its presence in the bronchial lumen indicates a very defined immunological affectation of the airway. This anatomopathologic alteration, in our opinion, guarantees the presentation of a persistent affectation of the tracheobronchial glandular physiology and of the airway. It also affirms that the bronchial localization is yet another of the anatomical foci to be considered in the diagnostic approach of pSS, given its good accessibility with induced sputum and its probable correlation with the lymphocytosis present in the bronchoalveolar lavage.3,6 It may be a valid diagnostic alternative to waiting to confirm pSS with a compatible biopsy of the labial mucosa, or rather, the verification of a more extensive or severe syndrome. Moreover, its presence in the airway is directed along the same critical line as the definition of SS postulated by Ramos-Casals et al.2 Given contemplation, this could contribute to changing the denomination of Sjögren's syndrome to that of Sjögren's systemic disease if persistent lymphocytosis is demonstrated at the bronchial level in patients with dry syndrome.2 For this reason, we consider that lymphocytosis is a relevant finding that is important to verify in patients with pSS and that it could possibly contribute to improving the classification of the immunological severity of the disease. Thus, the analysis of induced sputum can be considered a useful tool with a significant role in the evaluation of pSS patients.

On the other hand, it should also be noted that the majority of the sputa with lymphocytosis analyzed were accompanied by lipophages. This fact can be explained by the physiological affectation of the gastrointestinal mucosa by the disease itself, which would favor the microaspiration of gastric content to the bronchial Tree,6 or rather it simply reflects a greater degradation of the local glandular cell destruction. In our study, the presence of gastroesophageal reflux could not be confirmed by esophageal pH, although this fact has already been published in pSS.6

In this study, the following limitations can be considered. First of all, although it could be argued that there was a systematic appearance of lymphocytosis in the bronchial sample after nebulization depending on the induction time, this would equally affect all the subjects who underwent the same induction protocol and should have been found in all the subjects and not in just some. Nevertheless, although very improbable, it is not possible to completely rule out a partial recruitment of lymphocytes depending on the time and as an effect of the induction.24

In our study, we could not control the effect of systemic or inhaled medication that the patients were receiving, as it was very heterogeneous, therefore its influence on the interpretation of the duration of cough, lymphocytosis in sputum and methacholine was not evaluated.

In conclusion, respiratory symptoms (clearing of the throat, cough and dyspnea) are frequent in pSS, although their relationship with bronchial hyperresponsiveness and airway inflammation is variable. The lymphocytosis of the airway is the most constant pathological inflammatory finding, which probably defines, more than a syndrome, Sjögren's systemic disease respiratory affectation. Induced sputum is a useful tool in the evaluation of the bronchial affectation of pSS.

The authors declare that they have no conflict of interest.