Mushroom compost worker's lung has a prime suspect — and it is not a fungus

A decade of admissions from one compost plant

Between 2009 and 2019, twenty men employed at a mushroom compost plant in eastern Poland were admitted to the Clinic of Pneumonology, Oncology and Allergology at the Medical University of Lublin with symptoms that appeared when they handled compost dust. Eighteen were diagnosed with hypersensitivity pneumonitis; the other two with organic dust toxic syndrome, a separate condition that lacks the specific allergic reaction and leaves the imaging clear. These are not twenty cases of one illness. The case series, first published online on 18 June 2026 in Annals of Agricultural and Environmental Medicine (issue 2/2026, volume 33, pages 183–191), names the most probable agent of those symptoms, and it is a bacterium rather than the mushroom or its spores: Thermobifida fusca, a thermophilic actinobacterium that grows in self-heating organic matter.

The patients were aged 23 to 55, median 33. Cough was reported by 19 of the 20, breathlessness by 18 and chest tightness by 14, and auscultation found inspiratory crackles in 16. Routine lung-function testing, by contrast, came back normal in 16 of the 20 — in this series it mostly failed to register the disease at all. High-resolution CT showed ground-glass attenuation alone in 15, the acute form, and ground glass with fibrosis in three, the subacute form; the remaining two showed nothing. The illness has been called mushroom compost worker's lung since Phillips and colleagues proposed the name in 1987, and this team adopts it because most reported cases have occurred not among pickers but among the people who make the growing medium.

Narrowing two bacteria down to one suspect

Two thermophilic actinobacteria — bacteria, not fungi — were cultured from the compost the men had worked with. Thermobifida fusca strain X-1 was present at 2.45 million colony-forming units per gram and Laceyella sacchari strain X-2 at 0.1 million, nearly 25 times fewer. The two were told apart by colony form, by microscopy and by chemotaxonomy, and the species assignment rests on a short 16S rRNA fragment rather than a genome: 320 base pairs for Th. fusca and 471 for L. sacchari, lodged with GenBank on 14 February 2025 and released five days later. Those records give mushroom compost as the isolation source, Poland as the country and 2017 as the year of collection.

The immunological work was done on six of the twenty patients — numbers 4, 5, 12, 13, 14 and 15, chosen for the severity of their symptoms — and that denominator matters, because the abstract leaves it out. Three of those six had strong or very strong precipitins to Th. fusca antigen and none reacted to L. sacchari; the positive reactions appeared only in men who had done the job for between 3.5 and 11.5 years, and in none of those with a month to a year of exposure. The same six then inhaled a very small dose of the antigen, about 2.16 µg absorbed over twenty breaths, after a control test with saline. Four responded with a fall of at least a tenth in forced vital capacity or FEV1 at one or more of the measurement points — clearest eight hours after the challenge, though in some men it came at four hours or at twenty-four. Three coughed and one became breathless.

What a case series cannot settle

The finding is not that bacteria rather than fungi are to blame; that was already the accepted view. Reviews of hypersensitivity pneumonitis have long listed thermophilic actinobacteria as the agents of mushroom worker's lung, and Thermobifida was among the commonest actinobacteria in Hungarian mushroom compost profiled by molecular methods, although proteobacteria prevailed in that compost overall; the genus was recorded in Korean compost too, where other genera were commoner. What is new is the attribution to a species, and the authors word it as carefully as the evidence allows: Th. fusca is most probably a major causative agent, while the possible role of L. sacchari and of other thermophilic actinomycetes still has to be taken into account.

Nor is this a portrait of an industry. All twenty men came from a single plant, which the paper does not name, and no regulatory action is reported. There is no epidemiological figure either: no incidence, no prevalence, no workforce total against which twenty cases could be set. The Conclusions nevertheless state that the results confirm the common occurrence of the disease in Poland — that is the authors' assertion, and nothing in a twenty-patient series from one clinic can measure it. There is no limitations section and no conflict-of-interest statement, so each limit set out here has been drawn from the data rather than flagged by the researchers.

The Discussion closes by extending the risk to waste-handling workers, to farm workers exposed to self-heating heaps of hay and manure, and to biotechnology staff producing the enzymes this bacterium makes. None of those groups was studied, so that passage is inference, not result. And 'not a fungus' applies to this disease rather than to fungal illness in general: the same paper notes that Agaricus bisporus spores, while not implicated in this condition, have been reported as a rare cause of IgE-mediated allergy — occupational asthma, and anaphylaxis after the mushroom is eaten.

Where the mechanism becomes visible

The relevant part of the process can be seen from the yard road. A compost heap warms itself, and as the temperature climbs the community inside shifts towards organisms that thrive in the heat. Th. fusca is at home there: it develops in all manner of self-heating organic material, secretes cellulases and lignocellulases, and is a major degrader of plant cell wall material. Steam drifting off a heap marks that thermophilic phase, but it identifies nothing — which organisms are present is settled by culture and laboratory identification, never by looking.

For scale, the paper puts world production of A. bisporus at nearly 12 million tonnes on 2019 data, close to 9 million tonnes of it in China, with Poland fourth at around 0.36 million tonnes a year. The preventive measures it proposes are the authors' own suggestions rather than results: ventilation able to trap the often submicron particles of actinobacteria, efficient and comfortable respirators or masks, and, where the hazard grows, a bacteriocide safe for humans, such as propionic acid, added to the compost. What a reader can verify today is narrower and firmer: the two GenBank records, and the full paper, which is open under CC BY 4.0.