Langerhans cells proliferate within bronchial mucosa as well as in the peripheral lung. The cells have an ill-defined border, nuclei are convoluted often elongated, chromatin is vesicular, and nucleoli are small. In bronchi this proliferation causes necrosis of the mucosa, occlusion of the lumen, and finally scar tissue [4] (Fig. 7.1). Langerhans cell proliferation is accompanied by an infiltration of eosinophils, hence the old name eosinophilic granuloma (Fig. 7.2). These eosinophils are attracted by cytokines such as interleukin 4 secreted by the Langerhans cells (LH cells) [5]. Eosinophilic granulocytes might be the main cause of cytotoxicity releasing eosinophilic basic proteins and destroy the epithelium (Fig. 7.3). The granulomas undergo regression especially in patients with smoking cessation. The resulting scar has a stellate-like appearance and is surrounded by bronchiolectasis and emphysema blebs (Fig. 7.4). On CT scan this results in a characteristic picture called starry sky, where the dense scar shows tiny extensions (Fig. 7.5).

An underlying genetic abnormality, i.e., mutation in the BRAF gene, has recently been identified [6–8]. These mutations resulted in discussions, if LHCH should be regarded as a tumor. However, several issues remain to be solved, before this view can be accepted: lung cases are most often induced by smoking and in many patients will undergo regression in case of smoking cessation, which is unlikely in a tumor [3]. BRAF mutations are found in a minority of cases. On the contrary there exists a tumorous form, characterized by a multi-organ involvement, seen in children and young adults and not related to smoking [9, 10]. So probably we are confronted with two different diseases. Morphologically the reactive form cannot be discerned from the tumor form other than by involvement of at least two organ systems [11, 12]. Further investigations hopefully will increase our knowledge about this disease.

LH cells are part of the antigen-presenting cell system. Inhaled antigens are presented to LH cells and are taken up and processed by specific mechanisms involving Toll receptors and Langerin, a molecule with C-type lectin domain [13–15]. LH cells can interact with the specific as well as the innate immune system and play a role in control of infections as well as in the pathogenesis of asthma [16–18].

In the differential diagnosis, LHCH has to be separated from other histiocytosis or reticulum cell proliferations by their positive staining for CD1a and Langerin [19, 20] (Fig. 7.6), whereas the positivity for S100 protein is not specific. On electron microscopy the characteristic feature are the Birbeck granules, which resemble a tennis racket (Fig. 7.7). LHCH might be also diagnosed in samples of bronchoalveolar lavage: on cytology, usually macrophages and lymphocytes can be seen, but also increased numbers of eosinophils. By immunocytology using antibodies for either Langerin or CD1a, the number of positive cells should be >6/HPF; at least six field should be counted and a mean established (Fig. 7.8). Other histiocytoses need to be separated from LHCH. The histiocytes in Erdheim-Chester disease are negative for Langerin and CD1a, although BRAF mutations can be found as well in this disease. Systemic LHCH cannot be differentiated from the reactive pulmonary form as the cells express the same markers. Probably the proof of JL1 epitope of CD43 might help in this respect, as this epitope is expressed only on immature LH cells [21]. Histiocytic sarcomas are easier to separate, as they will present with nuclear atypia, increased mitosis, and invasive growth. However, it should be reminded that Langerhans cell sarcoma although rare does exist, showing the same marker expression as LHCH (see chapter on tumor).

An accumulation of alveolar macrophages within respiratory bronchioles and the adjacent centrilobular region of the lung lobules characterize RBILD (Figs. 7.10 and 7.11). The macrophages usually contain dirty brownish-yellow fine granular pigment (Fig. 7.12). Ultrastructurally this pigment represents phagolysosomes filled with tobacco waste including also some metal oxides such as cadmium oxide [24–26]. Functionally this macrophage accumulation obstructs the terminal bronchioles and impairs airflow, resulting in distension of alveoli and eventually rupture of septa.

Some authors use the diagnosis of RB only in those cases, where no other pathology is present; others this author included follow the morphology and accept RB diagnosis also in heavy smokers with lung carcinoma, in whom also RB is present. The argument is that RB is a disease of smokers, and there is no good argument why smokers are not allowed to have more than one disease, e.g., RB, carcinoma, LHCH, and emphysema. However, this results in different statistical figures: if RB diagnosis is only accepted presenting as a singular disease, it is rare; if diagnosed by its morphological features regardless of other smoking-induced diseases, it is a common disease, present in many patients with lung cancer. RB and RBILD in my opinion are subsequent stages of the same disease. In early stages accumulation of alveolar macrophages are concentrated within bronchioles. If tobacco smoke exposure goes on, more and more areas of the centroacinar region of alveoli are occupied by these macrophages, resulting in radiological ground-glass opacities, clinically called RBILD. Whereas the clinical symptoms of RB might be mild, the symptoms in RBILD are much more severe.

The mechanism of RB/RBILD is still not understood. However, there exists an experimental condition, which shows similar features: in previous experimental inhalation, study investigators used titanium oxide as an inert nontoxic control substance. In these studies it was shown that increasing the dosage of TiO₂ does cause a toxic reaction within the lung by an accumulation of macrophages resulting in obstruction of small bronchi and bronchioles and extension of the infiltrates into the centrilobular lung areas. Macrophages by phagocytosing TiO₂ released cytotoxic enzymes, which subsequently destroyed the wall of bronchi, bronchioles, and alveolar septa and finally resulted in scar formation and fibrosis of alveolar septa. This phenomenon was called overload [27, 28]. Morphologically it resembles RBILD. So probably RBILD is caused by an overload with toxic tobacco smoke products resulting in this accumulation of macrophages. However, it should be reminded that additional factors are acting, since RBILD is not seen in every smoker.

By definition DIP is characterized by an accumulation of pigmented smoker macrophages within alveoli, completely obscuring the peripheral airspaces. No infiltration of bronchioles is present (Figs. 7.13 and 7.14). Fibrosis of alveolar septa if present is mild. DIP can radiologically simulate a tumor with ground-glass opacity, not uncommonly misdiagnosed as adenocarcinoma in situ [22]. Only by immunohistochemistry using anti-CD68 antibodies the nature of the accumulating cells is becoming clear (Fig. 7.15).

For several decades, the discussion, if RBILD could be the early form of DIP or vice versa, is unsolved. But there are some aspects, which clearly separates both entities: DIP is a rare, whereas RBILD a common disease. Respiratory bronchiolitis is not seen in DIP and extension of macrophage accumulation beyond the centrilobular region is not seen in RBILD. If one disease arises from the other entities, some overlap features should be present. So I think this debate can be closed, just on logical aspects.

The reason of DIP might be explained similar to RBILD; however, as DIP is rare, there might be some other underlying disease modifiers, which we do not know. A few other causes have been reported in the literature such as steel welding fumes [36], waterproofing sprays [37], and toxin inhalation, as well as certain drugs [35].

Bronchoalveolar lavage can be of help in the diagnosis of DIP and RBILD. However, in both diseases pigmented alveolar macrophages dominate the cytological findings. Alveolar macrophages can be seen in up to 85 % of the total cells. Staining for hemosiderin can be positive in these macrophages; however, this is a mild finely granular staining, not coarse granular as in previous episodes of hemorrhage.