Hypersensitivity pneumonitis and metalworking fluids contaminated by mycobacteria

To the Editors:

We read with interest the article published by Tillie-Leblond et al. [1] relating to hypersensitivity pneumonitis (HP) in French automobile workers exposed to metalworking fluids (MWFs). Our group was involved in the UK outbreak investigation referenced in their article [2, 3], and have a clinical and research interest in this area.

Whilst Tillie-Leblond et al. [1] are correct in stating that the majority of MWF-HP outbreaks have occurred in the USA, the UK Powertrain and French outbreaks are not the only ones to have occurred in Europe. We have provided scientific support to three other similar outbreaks in the UK, all with confirmed cases of MWF-HP [4, 5]. In addition, we have diagnosed single cases of MWF-HP or asthma in workers from at least five other workplaces close to our occupational lung disease clinic. We are also aware of a published respiratory outbreak from Croatia [6], and have recently had separate personal communications with groups in Germany and Sweden relating to investigations of possible MWF outbreaks. Supportive evidence for a more widespread European problem come from cross-sectional studies that have demonstrated an excess of respiratory symptoms and asthma in machine shop workers in southern Finland [7], and an excess of wheeze, chronic bronchitis, chronic rhinitis and eye irritation in Swedish metalworkers [8].

It is clear, therefore, that this remains an important area of research relevant to European MWF-exposed workers, as the exact aetiology of MWF-HP has remained elusive [9] since Bernstein et al. [10] described the first cases in 1995. Although Mycobacterium immunogenum have been implicated as the cause in the French outbreak and a number of outbreaks in the USA, there is strong evidence against this being the cause of the UK outbreaks [3, 5]. In the UK Powertrain investigation, it was not possible to culture any opportunuistic mycobacteria or find any evidence of mycobacterial DNA by PCR in 125 MWF samples. Analysis of MWF samples from two other workplaces associated with UK MWF-HP outbreaks also found no detectable mycobacterial DNA. In addition to this, no demonstrable precipitin responses to extracts of M. immunogenum, Mycobacterium chelonae or Mycobacterium fortuitum were seen in 129 Powertrain UK workers, a group that included 17 cases of MWF-HP, 70 cases of occupational asthma, and 42 asymptomatic exposed controls. Further evidence against a mycobacterial cause came from specific challenges performed in two workers, where positive responses were seen after controlled exposure to used MWFs that did not contain mycobacteria [3].

Although referenced by Tillie-Leblond et al. [1], the detailed immunological investigation performed on workers from a MWF-HP outbreak in the USA, where mycobacterial contamination was identified [11], is not discussed in any detail. In this key study [11], in vitro secretion of interleukin-8, tumour necrosis factor-α and interferon-γ were measured in whole blood and from peripheral blood mononuclear cells in response to incubation with M. immunogenum antigen. These measures of immunity to M. immunogenum were able to distinguish between MWF-exposed and -unexposed workers, but not between workers with and without MWF-HP. This study serves to highlight the difficulties of interpreting immunological findings in HP, as many workers with demonstrable immune responses are asymptomatic, and never go on to develop disease [12].

Tillie-Leblond et al. [1] state in their article that the presence of five arcs to M. immunogenum by electrosyneresis differentiates MWF-HP from healthy exposed workers. Whilst this may have been true in most cases, it was not true for all, and may therefore be of limited clinical value in isolation. It should be noted that the two groups in this study had also been differentiated in other ways, based on the presence of symptoms, lung crackles and normal gas transfer prior to any immunological comparison. Whilst we agree with the authors that this test threshold needs further validation, there seems little point in aiming to differentiate those with MWF-HP from healthy workers without symptoms, as this can be done more easily by questionnaire. Whether the five-arc threshold has any diagnostic value is dependant on whether it can assist in differentiating workers experiencing work-related symptoms due to allergic occupational lung disease (HP and asthma) from those with similar symptoms due to chronic bronchitis, exacerbations of existing respiratory conditions, nonspecific irritant responses or humidifier fever. Alternatively, the utility of the test in preventing disease, as part of a prospective study of health surveillance linked to exposure modification in MWF apprentices, would also be of great interest. It is these areas that we believe merit further study by the authors to continue to improve the knowledge base in the complex area of MWF-HP outbreaks.

• ©ERS 2011