Pulmonary nocardiosis (PN) is the most frequent from of presentation of infection due to Nocardia spp. in humans.1 Its main risk factor is cellular immunosuppression, which mainly affects patients with hematologic diseases, solid organ transplant, those infected with human immunodeficiency virus (HIV) and patients who receive prolonged treatments with corticosteroids, cytotoxic drugs and/or biological therapies with anti-TNF.1–5 Its associated mortality is very high, ranging from 14% to 40%,1 increasing to 60%–100% in cases in which there is dissemination to the central nervous system (CNS).1,6,7

In recent years, there has been evidence of an increase in the number of cases of PN in patients with Chronic Obstructive Pulmonary Disease (COPD). An initial study by Menéndez et al. that evaluated 10 patients affected by PN showed that COPD was an important risk factor for developing the disease, even greater than having a neoplastic disease or being an HIV carrier.6 Later studies confirmed this observation. In 31 cases of PN,2 Martínez et al. observed that COPD was the third most frequent risk factor, surpassed only by chronic steroid treatment and solid organ transplantation. Likewise, in addition to confirming that lung affectation was the most frequent form of presentation of Nocardiosis in humans, other recent studies3,6 observed that the most prevalent underlying conditions were HIV infection3 and the presence of chronic respiratory diseases like COPD.6,8

All these findings suggest that COPD, especially in advanced stages, can play a very important role in the pathogeny of PN. COPD is a disease that has a great impact in our setting, affecting almost 10% of the population between the ages of 40 and 70, and it is already the third cause of hospitalization and the fourth overall cause of mortality. This mortality rate increases year after year, and it is predicted to become the third cause of death by the year 2020.9,10 Due to this reason, it is expected that PN in COPD patients will be seen more and more frequently and will acquire great importance due to its potential impact on the prognosis of the disease.

To date, no studies have evaluated either the factors that associate PN in COPD patients or the prognosis of the disease in this group of patients. The objective of the present study is to describe the characteristics of COPD patients who present with PN and to know those factors that are associated with its prognosis.

A retrospective study including all the COPD patients diagnosed with PN at Hospital de la Santa Creu i Sant Pau in Barcelona from 1997 to 2009.

The present study describes the characteristics of the patients with COPD who presented with PN and the factors that are associated with a poor prognosis of the disease. Until now, such data have not been published, and given the high prevalence of COPD and the potential impact of PN, they may greatly influence daily clinical practice.

The most important findings are that PN in COPD patients mainly affects those with advanced lung disease who receive systemic corticosteroid treatment, that its associated mortality is very high and that the presence of active neoplasm, antibiotic treatment for less than 3 months and systemic corticosteroid therapy are factors that are related with mortality 12 months after the diagnosis of PN.

The characteristics of the COPD patients studied demonstrate some important differences when compared with data from non-COPD patients with PN that had been previously reported. The most notable characteristic of our group is that it affects patients with moderate-intense lung function alteration, with the presence of chronic respiratory failure in more than half of the patients. To date, no other study had analyzed lung function in patients with PN. Another important finding is that more than half of the patients had received systemic corticosteroid treatment prior to diagnosis. This datum coincides with previous studies that related steroid treatment with the presence of PN in different patient populations.1,2,6,14,16

From the rest of the variables studied, we have observed a marked difference in the incidence between sexes, with a high predominance among men, which is probably explainable by the lower prevalence among women of smoking history and COPD diagnosis.9,10 It is also striking that less than 5% of the patients present a high number of associated comorbidities (4 or more), reinforcing the finding that PN is more associated with advanced pulmonary pathology and not a more general multi-pathological state.

From a clinical standpoint, the presentation of PN in COPD patients is very unspecific. Cough and dyspnea are the most frequent symptoms, and there is a lack of fever in more than half of the cases, unlike in other PN patient populations where studies have shown that fever was a frequent symptom.1,2,17 Likewise, the presence of COPD may favor a predominance of respiratory symptoms over systemic ones, a fact that was confirmed in a former study of 10 PN cases where cough was also the most frequent symptom in patients with chronic pneumopathies.6

In this group of patients, no specific radiological data for the presence of PN have been found. The most common form of presentation was alveolar infiltrates, coinciding with the study by Martínez et al.2 Cavitation, the most characteristic radiological finding reported in previous studies in immunosuppressed patients and PN,18,19 was only found in approximately 10% of the COPD patients.

As for microbiology, the species of Nocardia that was most frequently isolated associated with COPD was N. cyriacigeorgica (56.7%). This contrasts with previous studies in other patient groups where the most frequently isolated species were Nocardia asteroides,2,20,21Nocardia farnicia8 and Nocardia nova complex,22 although these results are difficult to compare due to the recent taxonomic changes in the genus Nocardia.13,23

The overall mortality one year after diagnosis with PN was 33%, which is considerably high when compared with other previous studies of COPD patients who presented exacerbations requiring hospitalization, where mortalities of 22%24 and 23%25 were detected 12 months after the exacerbation.

From a prognostic standpoint, three factors have been related with mortality one year after diagnosis. First of all, the presence of active neoplasm, a fact that was expected and can justify both an altered baseline immunological state as well as being a direct cause of demise, a fact that has not been analyzed in this present study. The second factor is the presence of treatment with systemic corticosteroids prior to the diagnosis of the infection. It is well known that corticosteroids, through their anti-inflammatory genomic mechanism of action,26 modulate the inflammatory response of the host against germs, and that its chronic administration may be the cause of immunosuppression and associated severe infections,27 a circumstance that could explain this finding.

Finally, it has been reported that antibiotic treatment directed at Nocardia administered for less than 3 months is associated with greater one-year mortality after diagnosis. The current clinical practice guidelines about the treatment of the disease28 are not clear in these cases. They recommend treatment for 12 months when there is affectation of the CNS and immunosuppression and for 3 months in the cases of cutaneous nocardiosis, but there is no clear indication for these patients. This finding can reinforce the need for prolonged treatment in cases of lung affectation with Nocardia spp., regardless of the immunological state of the patient and whether the affectation is only bronchial or parenchymatous.

It is important to underline that, in most prior studies about PN,1,6–8 the factor that was most related with mortality was dissemination of Nocardia to the CNS, a fact that occurred in only one patient in our study group. This demonstrates once again that the behavior of Nocardia spp. in patients with COPD differs from its behavior in other risk groups.

It should be emphasized that this present study has some limitations that should be considered. In the first place, it is a retrospective study in which many potential biases have not been controlled. The number of patients studied is too small to be able to perform a multivariate analysis, correct for multiple variables or reach more solid conclusions. This is despite the fact that it is a very similar number to all the previously published studies about PN,3,6 and that it is the most numerous study with patients with PN and lung pathology published to date. One of the main factors that have limited the inclusion of patients has been the required spirometry for the diagnosis of COPD, as indicated in current clinical practice guidelines.11 Another limitation is that mortality was studied globally, not only mortality attributable to PN; thus, in some cases mortality may have been due to pathologies unrelated with PN. Finally, other important limitations of the study are that we do not know the exact dosage of corticosteroids administered or the stage of the active neoplasms, which are factors related with mortality. In addition, we lacked chest CT for all the patients to better understand the presence or absence of alterations in the pulmonary parenchyma.

The main conclusions of this study are that PN is a severe disease with a high short- and mid-term mortality rate in COPD patients. Its presentation is varied and especially affects patients with moderate-intense alteration of the lung function, chronic respiratory failure and recent systemic corticosteroid treatment. The presence of active neoplasm, recent systemic corticosteroid treatment and antibiotic treatment during a period of less than 3 months are the factors associated with poor prognosis and with mortality 12 months after the diagnosis of PN. Although extensive prospective studies are necessary to validate these findings, these results may promote the active, early identification of the bacteria in this type of patients, while trying to avoid adjuvant corticosteroid treatment as well as maintaining antibiotic treatment aimed at Nocardia for more than 3 months in order to improve prognosis.

The authors declare no conflict of interest.