INTRODUCTION

TOP ABSTRACT INTRODUCTION CASE REPORT DISCUSSION REFERENCES

Nontuberculous mycobacteria (NTM) are a heterogeneous group of organisms that occasionally are a primary cause of lung disease but more commonly affect patients with underlying chronic lung disease such as bronchiectasis, pneumoconiosis, or healed tuberculosis (1, 2). From an epidemiologic perspective, patients with cystic fibrosis (CF) are at high risk of NTM infection, but the clinical signs and symptoms of NTM are often difficult to distinguish from chronic bacterial pulmonary infections found in the advanced stages of CF (3). This may result in an underestimation of the contribution of NTM to the decline of lung function in cystic fibrosis. With the increased survival of patients with CF and with the improved surveillance and microbiologic methods of detection, the prevalence of mycobacterial colonization among the CF population appears to be increasing (6). The significance of this is not yet firmly established, and distinguishing between the carrier state and active disease remains difficult. Although our understanding of the role of NTM in CF is increasing with the passage of time, there are many unanswered questions. We report a case of a patient with cystic fibrosis who developed active disease with Mycobacterium abscessus. This case highlights a number of important points regarding the treatment and management of a patient with CF and mycobacterial disease. It also suggests that invasive disease may be present even if the patient is relatively asymptomatic and that the concept of chronic endobronchial colonization should not always be assumed.

	CASE REPORT

TOP ABSTRACT INTRODUCTION CASE REPORT DISCUSSION REFERENCES

Our patient was a 33-yr-old woman in whom CF was diagnosed in 1985 when she was 19 yr of age and had presented with hemoptysis and a history of recurrent sinopulmonary infections. A chest radiograph at the time revealed bronchiectasis, mainly of the upper lobes, and nodular densities in the right lower lobe. She had two bronchoscopies, both of which confirmed the presence of a nontuberculous mycobacterium identified as Mycobacterium fortuitum complex on culture. She underwent an open lung biopsy. At surgery, a diffuse nodularity of the lungs, mainly involving the lower lobes, was noted. An incisional biopsy of the posterior segment of the right lower lobe was performed. Histopathologic examination of the specimen showed granulomatous inflammation with focal necrosis principally involving terminal bronchioles (Figure 1). Occasional compact granuloma of the tuberculoid type with Langhans' giant cells were present (Figure 2). Small tubercules also lay in the perivascular adventitia (Figure 3). A rare acid-fast bacillus was seen in the cytoplasm of histiocytes. The findings indicated mycobacterial infection in an early bronchocentric phase, but caseous necrosis characteristic of established mycobacterial infection was not present. The lung culture failed to yield any mycobacteria but Actinomyces meyeri was isolated on an anaerobic culture. A subsequent sweat test was positive (sweat chloride, 99 mEq/L) and a recent CF mutational analysis revealed a genotype of dF508/ 2789+G  A. She was treated with high dose penicillin, oxacillin, and tobramycin. Because of her clinical improvement, it was decided to not institute therapy for mycobacterial disease at that stage. Baseline PFTs on discharge were FVC, 96% and FEV₁, 87%.

View larger version (182K): [in this window] [in a new window]   Figure 1.   Lung biopsy with granulomatous inflammation filling a terminal bronchiole and destroying portions of the wall (hematoxylin-eosin stain; magnification: ×10). A segment of intact respiratory epithelium remains (arrows). Necrosis involves the center of the bronchiolar inflammation (N).

View larger version (170K): [in this window] [in a new window]   Figure 2.   Two compact tuberculoid granulomas each having a multinucleated Langhans' histiocyte at the center (L) (hematoxylin-eosin stain; magnification: ×20).

View larger version (155K): [in this window] [in a new window]   Figure 3.   Small compact granuloma (arrow) lies in the adventitia surrounding an artery (A) away from the main mass of inflammation. A small bronchiole also contains histiocytes (B). (hematoxylin-eosin stain; magnification: ×20).

For the next 11 yr, she did well, requiring only three courses of intravenous antibiotics. Her medications included oral bactrim, dornase alpha (from July 1993), and vitamins, and she was pancreatic sufficient with normal growth and stable weight. Repeated sputum cultures inconsistently yielded Pseudomonas aeruginosa, Staphylococcus aureus, Candida, and Aspergillus but persistently yielded NTM (identified subsequently as M. chelonae, which based on susceptibility to cefoxitin was later reclassified as M. abscessus). A total of nine sputum samples were positive for NTM during this period. Only one sputum sample was smear positive (1 to 10 AFB per field) and all yielded 10 to 50 colonies of the Mycobacterium on culture. PFTs showed a slow decline of approximately 2% per year (Figure 4), and her CXR appearance showed minimal interval change. In an effort to slow the progession of her disease, high dose ibuprofen (1,600 mg twice a day) was begun in February 1996. Three months later, she was admitted with increased cough, low grade fevers, decreased energy, a right upper lobe infiltrate, and a 15% decline in her PFTs (FVC, 82% and FEV₁, 56%). Once again, her sputum was positive only for M. chelonae. Despite a 2-wk course of antipseudomonal antibiotics with improvement in her CXR appearance, her symptoms recurred soon after discharge and she continued to complain of intermittent low grade fevers, a decrease in stamina, and an increase in her baseline cough. A PPD was negative and control was positive. These symptoms persisted despite two subsequent courses of oral antistaphylococcal antibiotics and one course of inhaled tobramycin. Her PFTs were persistently below baseline (FVC, 81% and FEV₁, 56%). Because of the concern of active mycobacterial infection, a high resolution computerized tomography (HRCT) was performed in November 1996. This demonstrated moderate diffuse bronchiectasis with bronchial wall thickening but no abscess formation and no nodules (Figure 5A). Bronchoscopy with bronchoalveolar lavage (BAL) confirmed the presence of S aureus, Aspergillus (skin test negative, IgE level of 73 U/ml), and M. chelonae. The BAL samples were smear negative, but greater than 50 colonies of M. chelonae were isolated on culture. A further course of treatment with dicloxacillin and itraconazole failed to produce a clinical improvement. She presented in February 1997 acutely ill with high spiking fevers. A CXR at this time showed bronchiectasis, but, in addition, a large 3.5 × 2.5 cm cavitary lesion with an air fluid level in the superior segment of the left lower lobe, which had not been present on the CT scan 3 mo earlier. This was confirmed on chest CT (Figure 5B). Bronchoscopy was repeated and copious purulent secretions were drained from the left upper lobe. On direct smear, these secretions were filled with acid-fast bacilli. The cultures yielded only M. chelonae. On the basis of these findings and the escalating clinical course, she was begun on intravenous cefoxitin and amikacin and oral clarithromycin (500 mg twice a day), to which the organisms were sensitive, for a period of 6 mo with a dramatic improvement in clinical status and CT appearance. Amikacin levels were kept in the low therapeutic range (peak concentration, 15 to 25 µg/ml) because of tinnitus and progressive hearing loss. Ibuprofen therapy was discontinued, and a comprehensive evaluation for causes of altered immunity, including an immune work-up, a colonoscopy, and a gynecological examination, failed to identify any underlying condition. In an effort to achieve complete eradication of the organism, a surgical lobectomy was considered 3 mo into therapy. As her disease was not localized to the upper lobe alone, the consensus was that the loss of a lobe in this situation outweighed the potential benefit. A follow-up HRCT scan 10 mo after therapy showed complete resolution of the cavity (Figure 5C). Despite a dramatic clinical response and a full 6 mo of intravenous therapy, her sputum cultures continue to remain smear negative but culture positive. The patient was asymptomatic, with stable PFTs (FVC, 90% and FEV₁, 63%) for 13 mo when her clinical symptoms recurred with evidence of a further abscess on chest CT (Figure 5D). She was re-treated with a further 5 mo of intravenous cefoxitin and amikacin with oral clarithromycin in addition to nebulized amikacin. Therapy was discontinued after 5 mo because of a drop in her creatinine clearance from 78 ml/min/m² (normal range, 60 to 80 ml/min/m²) at the end of her first course of treatment to 58 ml/min/m². She has been maintained on nebulized amikacin for the past 2 mo and a recent culture for NTM was negative for the first time since commencing therapy.

View larger version (14K): [in this window] [in a new window]   Figure 4.   Graph of pulmonary function over time.

View larger version (127K): [in this window] [in a new window]   Figure 5.   Radiologic course of M. abscessus infection by serial CT scans. (A-C, cuts at same level; D, higher level). (A) November 1996. Bronchiectasis without any cavitation. (B) February 1997. Large cavity in left lower lobe associated with a clinical deterioration. (C ) June 1998. Resolution of cavitation 10 mo after completion of therapy. (D) October 1998. New cavity with worsening of clinical status.

	DISCUSSION

TOP ABSTRACT INTRODUCTION CASE REPORT DISCUSSION REFERENCES

This case report is important for a number of reasons. Firstly, it gives us hitherto unavailable insight into the natural history of M. abscessus in a patient with underlying lung disease and raises legitimate concerns about the use of immune-modulating drugs such as ibuprofen in these patients. Secondly, it addresses the difficulty faced by many clinicians in distinguishing between colonization and active mycobacterial disease and suggests that this distinction may in fact be arbitrary. Finally, it provides us with some useful information regarding the treatment of the disease and the likelihood of achieving meaningful eradication of the organism.

Mycobacterium abscessus is a member of the rapidly growing mycobacteria. With improvements in identification, it is now known that M. abscessus, which was formerly identified as M. chelonae subspecies abscessus and prior to that as part of the Mycobacterium fortuitum complex, accounts for approximately 80% of rapidly growing mycobacterial respiratory disease isolates (1). Prior to the 1990s, there were 18 reports of clinically important NTM infection complicating CF of which seven were cases of M. abscessus (10). In one case, M. abscessus infection was felt to be the direct cause of death. To date, of a total of 126 reported cases of positive mycobacterial cultures in patients with CF, 38 have been attributed to M. abscessus, making it the second most common isolate after M. avium complex (5). Although the natural history of this disease can vary, it appears to follow a more fulminant course in patients with underlying lung diseases such as cystic fibrosis (17). The concept of "colonization," defined as the isolation of the organism from the respiratory tract without evidence of tissue invasion, was supported by several investigators in the past, although mainly in association with M. avium complex infection (1, 18). With the emergence of high resolution CT scanning (HRCT), it now appears that many of these patients have features of parenchymal involvement, usually of a nodular pattern, suggesting that the term "colonization," at least in some patients, may be erroneous (19). In our patient, early histologic changes provide the most conclusive evidence that NTM, even in the absence of symptoms, cannot always be considered innocent colonizers of the respiratory tract in patients with CF. Her original biopsy showed necrotizing granulomas and granulomatous inflammation with rare mycobacteria in the necrotic areas indicative of active infection. Furthermore, granulomas away from necrosis are consistent with spread of organisms in the lung. Despite that, she remained clinically well and she did not experience a dramatic decline in pulmonary function for more than 10 yr. Without the benefit of serial tissue examinations, we cannot state conclusively that our patient had long-standing indolent infection as opposed to chronic endobronchial colonization with intermittent invasive disease, although we feel it is unlikely that the infection resolved without therapy during this time. Therefore, the repeated isolation of the same Mycobacterium from a patient with CF may be indicative of indolent infection and these patients may be at significant risk of developing clinical symptoms that will ultimately require treatment. This observation is supported by a recent study by Tomashefski and colleagues (20) who found that patients with CF who have multiple positive sputum cultures for NTM were more likely to have histologic evidence of granulomatous mycobacterial disease at post-mortem. It also suggests that infection may be subclinically active for a long period of time and that it may account for or contribute to a progressive decline in PFTs. Attempting to determine if disease is locally invasive in CF is difficult. Subtle changes on serial HRCT scanning may enable one to make an earlier diagnosis and to intervene prior to the development of cavitation and/or irreversible lung damage, but the progressive nature of the lung disease associated with cystic fibrosis substantially limits this capacity. In the future, rather than using terminology such as colonization and active disease, perhaps a more appropriate distinction would be between those patients who require immediate therapy on the basis of the rapid progression of their disease and those in whom the decision can be delayed (5).

Despite the fact that our patient had an underlying lung condition, her disease was indolent and only became clinically active after 13 yr. This protracted course of invasive M. abscessus in the context of CF has not been documented by histology previously in the literature. The rapidity of her ultimate deterioration did raise concerns about an underlying immune deficit or malignancy. The patient has never been treated with inhaled or systemic steroids, and immunologic tests including T- and B-cell subsets were normal. There was no evidence of an underlying malignancy. Of note was the fact that she had recently been started on ibuprofen, a nonsteroidal anti-inflammatory drug. High dose ibuprofen was found to halt the rate of progresion of mild pulmonary disease in patients with CF (21). It has been shown that ibuprofen, in high doses, inhibits the migration, adherence, swelling, and aggregation of neutrophils as well as the release of lysosomal enzymes (21, 22). It has also been shown to inhibit monocyte chemotaxis in vitro (23). The close association of the initiation of this new treatment with a sudden change in the clinical course suggests that this drug could have altered the immune milieu, allowing progession of infection. Ibuprofen has also been shown to have a direct effect on CFTR function by inhibiting chloride secretion (24). Whether ibuprofen, by some mechanism, could have had a suppresssive effect on the CFTR function making an individual more susceptible to mycobacterial invasion can only be speculated at this time. If, however, such a relationship exists, it could explain the temporal relationship between the commencement of ibuprofen in our patient and the progression of a long-standing quiescent NTM infection.

The treatment of M. abscessus is complex. The isolates are uniformly resistant to standard antituberculosis drugs and are generally susceptible only to parenteral antibiotics and the newer oral macrolides. The organism is notoriously difficult to eradicate. In a report of 154 patients infected with rapidly growing mycobacteria (17), only 10 patients infected with M. abscessus exhibited a "cure," the latter being defined as a return of respiratory symptoms to baseline and reversion of sputum to smear and culture negativity for at least 1 yr. Of these 10 patients, seven had received amikacin, cefoxitin, or imipenem for 1 to 3 mo followed by surgical excision, whereas only three subjects had been sucessfully treated with antibiotics alone for an unspecified duration. No documented instances of permanent sputum conversion to culture negativity and permanent symptomatic improvement occurred using parenteral antibiotics alone. Preliminary studies using monotherapy with oral macrolides would indicate that these drugs are also unlikely to result in a microbiologic cure (1). Surgical resection of localized disease is thought to be the only effective method in achieving complete eradication of the organism but very few patients with CF present with or have the disease diagnosed at a stage early enough to be amenable to surgical resection. With the belief that 6 mo of triple therapy could be potentially curative, a belief subsequently endorsed by the 1997 ATS Guidelines (1), it was elected to treat our patient with intraveneous amikacin and cefoxitin in addition to oral clarithromycin for that length of time. Although there was a dramatic clinical and radiologic improvement, including complete resolution of the cavity, her cultures remained positive even while receiving therapy, suggesting that complete eradication of the organism, even with this aggressive therapy, is an unachievable end point in patients with CF. In hindsight, this outcome was not surprising and supports the cumulative experience expressed in the CF literature to date that eradication of this organism with antimicrobial agents alone is unlikely. The decision to discontinue clarithromycin was based on the available evidence that triple therapy had failed to produce a microbiologic cure, making it unlikely that continued therapy with an oral macrolide would do so. In addition, there was concern about the potential ototoxicity and nephrotoxicity of this drug in a patient with tinnitus and hearing loss and prolonged exposure to other nephrotoxic drugs (25). Finally, there was concern that monotherapy could lead to the emergence of resistant strains. In retrospect, the reemergence of clinical disease within 13 mo of discontinuing therapy would lead one to speculate whether the continuation of suppressive therapy could have prevented the second relapse. Possible therapies for this patient in the future include intermittent suppressive therapy such as periodic parenteral antibiotics, oral macrolide therapy, or nebulized amikacin to control symptoms and progression of lung disease. Whether this approach will result in the emergence of resistant organisms is still unknown.

We have presented a patient with CF and a history of multiple positive sputum cultures for M. abscessus who developed clinically apparent mycobacterial disease 13 yr after the organism was first isolated. Early histologic evidence suggests that this organism did not simply colonize the respiratory tract but was in fact subclinically invasive for 13 yr prior to detection. Dramatic clinical and radiologic improvement was achieved with 6 mo of intravenous therapy, but without eradication of the organism, and clinical relapse occurred within 13 mo. The optimum treatment of M. abscessus in CF has yet to be determined and microbiologic cure may be an unrealisitic end point. Certainly 6 mo of triple antibiotic therapy was inadequate in this case and raises the question as to whether these patients may benefit from long-term suppressive therapy. Coincidentally, high dose ibuprofen had been recently started but the contribution this has had, if any, to the activation of the infection can only be speculated at this time.