Bronchial inflammation plays a major role in the pathogenesis of important respiratory tract diseases, such as asthma or chronic obstructive lung disease (COPD). A tool for evaluating the type and intensity of bronchial inflammation would be of great benefit in the assessment and control of these diseases, particularly in the more severe forms.

Endobronchial biopsy is the gold standard in the study of bronchial inflammation; being an invasive procedure, however, its use as a diagnostic tool is limited.1 Accordingly, interest is growing in non-invasive procedures, and new techniques for measuring fractional exhaled nitric oxide (FENO) and inflammatory cell count (ICC) in induced sputum (IS) are attracting particular attention.

While FENO identifies eosinophilic bronchial inflammation, it does not recognize other inflammatory types, so its utility is limited.2 IS is a validated standardized technique, considered the gold standard non-invasive methods for evaluating bronchial inflammation and for distinguishing between inflammatory phenotypes.3 Its clinical applications are becoming more refined, and it is of particular benefit in asthma, due to its high yield. It can be used as a complementary diagnostic tool in asthma,4 and is of benefit in determining inflammatory phenotypes,5 adjusting treatment, and predicting therapeutic response.6,7 It has also been indicated in the management of difficult-to-control asthma (DCA) in Spanish and international guidelines.8–10 Its use is not restricted purely to asthma: it is also useful for determining the etiology of chronic cough,11,12 gastroesophageal reflux (GER)13 and other entities, such as COPD, infectious diseases, eosinophilic bronchitis, lung cancer, interstitial lung diseases, and heart failure.14–19

However, despite its validity and applicability, ICC in IS is not routinely performed in standard clinical practice, probably because it requires a certain degree of technical experience in obtaining, manipulating and interpreting the samples, in addition to being laborious and costly. Nevertheless, the data provided by IS testing is so obviously of interest10 that we are surprised how rarely it is used in high-level pulmonology departments. The aim of this study was to evaluate the contribution of ICC in IS in healthcare practices.

A total of 171 patients were included, 95 (55.6%) of whom were women, and mean age was 50 (±15) years. Sociodemographic and clinical data of the study group are shown in Table 1.

Patients included in the study were classified according to the reason for ordering an IS. The most common reasons were asthma in general (103 [59.20%]) and DCA (34 [19.54%]), to complete the diagnostic process or to determine the inflammatory phenotype. The next most common reasons were chronic cough (19 [10.9%]) and suspected GER (15 [8.6%]). Three cases (1.7%) were classified as “other causes” and were excluded because of their heterogenic nature.

Patients were classified according to the severity of their asthma and DCA, as follows: intermittent asthma 10 (7.2%) cases; mild persistent asthma 41 (29.9%), moderate 33 (24.1%) and severe 45 (32.8%). Eight cases (5.8%) were being investigated for suspected asthma. The mean Asthma Control Test score in the asthma group was 21.36 (±5.10) and in the DCA group, it was 15.9 (±5.12).

Table 2 shows the characteristics of the IS samples. In the ICC, the highest percentage of eosinophils was observed in the DCA group, and the highest lipophage percentage was in patients with chronic cough and GER. IS was classified as “high quality” in 98 (57.3%) cases. The paucigranulocytic phenotype was predominant among the overall study cohort (88 [51.5%]) and among the subgroups for which IS was requested, although the eosinophilic phenotype emerged as the second most common phenotype in the asthma and DCA groups with 32 (31.1%) and 10 (29.4%) cases, respectively.

A weakly positive correlation was observed between age and neutrophilia in ICC (R=0.15, P=.05). However, no correlation was observed between the presence of eosinophils in IS and FENO values (R=0.12, P=.14).

Finally, the IS result contributed to clinical decision-making in 115 (67.3%) of cases, assisting diagnosis in 98 (57.3%) patients, and therapeutic decisions in 85 (49.7%). The results are listed by reason for IC in Table 3. IS was of greatest benefit in decision-making in the general asthma group, followed by DCA: 74 (71.8%) and 23 (67.6%), respectively, although there were no differences in its utility among the different reasons for request. In the asthma group, a similar contribution to the diagnostic process (59 [57.3%]) and to therapeutic decision-making (58 [56.3%] was also observed. However, in the DCA group, IS provided more useful information for diagnostic (23 [57.6%] patients) than therapeutic (15 [44.1%] patients) decision-making. Likewise, in the chronic cough group, IS was more useful for guiding diagnosis (Table 3). In patients with suspected GER, the identification of a significant number of lipophages supported the diagnosis in 8 (53.3%) patients.

The main finding of our study is that ICC in IS is useful in routine clinical practice, since it provides important information for clinical decision-making in two thirds (67.3%) of patients in whom this complementary procedure was requested.

Most experience and data on the use of IS are from asthma patients. Accordingly, IS was usually ordered in the context of asthma and DCA. Diagnostic applications are based mainly on the well-established relationships between asthma and eosinophilia in sputum.5 If a cutoff point is set at 1% eosinophils in sputum, IS has a sensitivity of 80% and a specificity of 95% for confirming a diagnosis of asthma.20 Although the most typical finding is eosinophilia in sputum, a systematic review of over 25 studies in asthma patients from different populations with varying degrees of severity,32 and more recently, a study which analyzed 508 patients with asthma,33 found that approximately 50% of these patients had non-eosinophilic asthma. In our series, more patients had non-eosinophilic asthma than generally described in the literature. However, the contribution of IS to confirming the diagnosis or determining the inflammatory phenotype was 67.6% in DCA and 57.3% in asthma.

Characterization of the inflammatory phenotype has important therapeutic implications, since each has a different response to anti-inflammatory treatment. Little et al.34 observed that patients with asthma who showed eosinophilia in IS had a significant increase in FEV1 after 2 weeks of treatment with SCS. Moreover, IS tests can reduce the number of exacerbations in asthmatic patients. Two randomized trials6,7 conducted in asthmatics whose treatment was adjusted according to clinical practice guideline recommendations or according to their IS eosinophil count showed a reduction in the number of exacerbations in patients whose treatment was established on the basis of their ICC. These results were confirmed in a recent systematic review,35 and are reflected in the latest ATS/ERS guidelines on the management of severe asthma.24 These recommendations state that in hospitals with the appropriate experience, treatment of patients with severe asthma should be based on both clinical criteria and monitoring of IS eosinophil levels. We also studied the contribution of IS to therapeutic decision-making in asthma patients. Results centered on the eosinophilic and neutrophilic phenotypes, but in some cases the decision to reduce steroid treatment was made in clinically stable patients who revealed a paucigranulocytic phenotype, according to clinical practice guidelines. Nevertheless, these premises only apply to selected patients, since the monitoring of eosinophilia in IS has not been adopted as a routine method in standard clinical practice.

In patients with chronic cough, defined as a >3-week history of cough not immediately preceded by an acute process, IS is recommended to complete the diagnostic process.11,12 Bronchial eosinophilia revealed in patients with chronic cough can help guide the diagnosis of asthma, eosinophilic bronchitis, or occupational asthma, and can also predict a favorable response to steroid treatment. In our study, none of our chronic cough patients had eosinophilia in IS. However, in 3 (15.8%) patients, neutrophilia predominated, suggesting infectious disease. The finding of lipophages in IS (>7%) is 90% sensitive and 89% specific for a diagnosis of GER.13 Lipophages (>7%) were found in 8 study patients investigated for chronic cough and for suspected GER, which explained 42.1% of causes of chronic cough, and supported 53.3% of suspected GER cases.

In line with the previous literature,21,36 we found evidence among our series of a trend toward increased bronchial neutrophilia as age advances. However, more studies are need to determine if these findings affect the diagnostic yield of IS in elderly patients.

In spite of all the reports in the literature describing the benefits of IS, only Moritz et al.,37 in a series of 151 IS samples, found the results to be clinically useful in standard clinical practice, because they modified medical treatment in 82 (55%) of cases. The contribution was even greater in asthmatic patients: steroid dosing was modified in 48 patients (64.7%). The results of our study are similar to those described by Moritz et al. However, in our opinion, the contribution of the procedure to the diagnostic process must be included if the real benefit is to be studied.

In spite of the limitations of the retrospective nature of this study (which, for example, in the case of asthma made it impossible to determine any improvements in future risk), the utility of the procedure was evaluated by observers who were independent from the treating physicians, in order to reduce subjectivity bias. Moreover, IS samples were induced, processed and evaluated according to a systematized protocol, by highly experienced technical staff, in compliance with the strict quality control standards of our laboratory. These precautions contributed to reducing potential variability among observers assessing ICC and standardizing sample collection and processing.

To conclude, our study demonstrates that ICC in IS is beneficial in clinical decision-making, particularly in 2 common entities, asthma and chronic cough. We provide arguments for recommending the use of this technique in clinical practice, at least in referral pulmonology departments, particularly those equipped with dedicated asthma units.

The authors declare that they have no conflict of interests.