Adult cystic fibrosis presenting with recurrent non-tuberculous mycobacterial infections

Abstract

In November, 1997, a previously healthy 54-year-old white man presented with productive cough, low-grade fever, and chills resistant to outpatient treatment with clarithromycin and cefpodoxime proxetil. CT of the chest showed consolidation of the left lower lobe and lingula. Spirometry was normal, and HIV serology was negative. Sputum showed acid-fast bacilli (AFB) and cultures grew Mycobacterium avium intracellulare. The patient was treated with rifampicin, ethambutol hydrochloride, and clarithromycin for 2 years. In January, 2002, he presented to our clinic with recurrent symptoms. Chest CT showed only linear densities in the left lower lobe. AFB smears were again positive. Cultures grew Mycobacterium chelonae and the patient was treated with doxycycline, clarithromycin, and ciprofl oxacin. Several months later a repeat AFB smear was positive, with cultures growing Mycobacterium abcessus and M avium intracellulare. He was admitted to our unit, and highresolution chest CT showed mild bilateral central bronchiectasis. Barium swallow showed a normal swallowing mechanism without gastro-oesophageal refl ux and no laryngeal penetration. Tests for antinuclear antibody and rheumatoid factor were negative. α-1 antitrypsin and immuno globulin concentrations were normal. He was discharged 10 days after admission. Despite repeated attempts, the patient had fathered no children, therefore a sweat chloride test was done as an outpatient; the concentration was high at 96 mmol/L (normal 40–60 mmol/L). Cystic fi brosis (CF) genotyping showed that the patient was heterozygous for two diff erent CF trans membrane conductance regulator gene alleles: δF508 and R117C. The patient was treated with ciprofl axacin, clarithromycin, and ethambutol for 3 years. When last seen in March, 2006, the patient had negligible sputum production, and AFB smears have been negative since 2004. Non-tuberculous mycobacteria (NTM) are free-living organisms that are ubiquitous in the environment. Occasional isolates of NTM were generally thought to be contaminants or colonisers until the second half of the 20th century. As the prevalence of Mycobacterium tuberculosis declined in developed nations and modern microbiological methods were developed, the importance of NTM in human disease became increasingly evident. By the early 1990s, laboratories in the USA informally reported that about two-thirds of recovered mycobacteria species were NTM. Some of this increase is related to the high rate of disseminated M avium complex disease in patients with AIDS; however, there could also have been an increase in the rate of NTM infections in other groups of patients, such as those with CF. The classic description of NTM pulmonary infection is that of a patient with cavitary infi ltrates and underlying lung disease. We have seen two patients in their mid-50s, who developed recurrent NTM pulmonary infections and were sub sequently diagnosed with CF. Our second patient had congenital absence of the vas deferens and normal sweat chloride concentration; he was confi rmed to have CF by genotyping showing the δ508 and D1152H alleles. Among adults with CF, 63% are aged 18–29 years, 25% 30–39 years, 10% 40–49 years, and 2% over 50 years ; adults will represent >40% of the US cystic fi brosis population in 2005. Among the 22 301 patients in the 2000 CF Foundation registry, the diagnosis was established after the age of 17 years in only 831 (3·7%); sweat chloride tests were abnormal in 90% of cases. Although NTM infections in CF are well described, recurrent NTM infection as the initial presentation of CF has not been described. The reported prevalence of NTM in CF ranges from 3% to 30%. Contamination of mycobacterial cultures (especially by Pseudomonas aeruginosa) can lead to underestimation of the prevalence of NTM infection in CF. Controversy surrounds the distinction between NTM infection and simple colonisation in CF. Symptomatic patients with persistently positive AFB smears and new radiographic abnormalities are most likely to be truly infected. Patients with negative AFB smears but positive cultures with low numbers of colonyforming units are probably colonised. We believe our patients were actively infected and not just colonised because they were symptomatic and had repeatedly positive AFB smears, high colony-forming unit counts on culture, and new radiographic changes. Both also improved with appropriate anti-NTM therapy. Our patients were pancreatic suffi cient and their respiratory symptoms were mild and delayed in comparison to more severe CF phenotypes. They remain free of sputum colonisation by pseudomonas at age 60 years (case history presented in detail) and 57 years. We postulate that the benign course of their illness is related to the fact that they are heterozygous for two diff erent CF gene alleles, the usual δF508 and R117C/ D1152H—the latter two are associated with milder disease. We conclude that in patients with recurrent NTM infections and new radiographic abnormalities, a diagnosis of CF should be considered irrespective of the patient’s age.