Morphological characterization of pathological lesions

In the FE-treated group, neoplastic and non-neoplastic lesions being pleural mesotheliomas (13/80, 16.3%), pleural plaques or nodules (6/80, 7.5%), and pleural fibrosis (51/80, 63.8%) were diagnosed (Tables 1 and 2).

Table 1 Incidence of pleural mesotheliomas and histotypeTable 2 Histomorphological evaluation of lung samples in the FFE-treated SD rat study. Incidence of preneoplastic pleural nodules or plaques and fibrosis

Among the 13 cases of mesotheliomas, 6 were observed in males and 7 in females. In males, the most prevalent histotype was the epithelioid one.

Their histological features included tubule-papillary, trabecular, or adenomatoid architectural patterns, lympho-histiocytoid cytological features, or myxoid stroma [16, 23]. Two cases out of the six males were identified as biphasic (or mixed) subtype showing both epithelioid and sarcomatous components. Finally, only one case was identified as sarcomatous/fibrous subtype, composed of elongated/spindle cells arranged in solid sheets or within a fibrous stroma. In females the biphasic (or mixed) subtype was the most prevalent (4/7 cases), followed by the sarcomatous/fibrous one (3/7 cases).

Finally, all of the non-neoplastic lesions examined were accompanied by reactive inflammation affecting pleura and lung parenchyma. Moreover, fibrosis was associated with the presence of FE-fibres. No lesions were observed in untreated controls.

Semi-quantitative assessment of lung fibrosis

Forty fields per animal were analysed and scored using an Ashcroft scale. The distribution of frequency of the scores for each field is summarized in Fig. 1 and Table 3. For each animal we then calculated a mean and a median Ashcroft score, and these summary measures were used to calculate the mean Ashcroft scores and SD by sex and treatment. These results are showed in Table 4.

Fig. 1

Frequency of Ashcroft scores assigned in each field by sex and treatment

Table 3 Frequency of Ashcroft Scores by sex and group (number of fields and percentage)Table 4 Average Ashcroft scores by sex and group of the mean and median Ashcroft scores per animal

Table 3 reports the absolute number of times each score was used and the percentage frequency. A descriptive analysis of all evaluated fields shows how, both in females and males, more than 2/3 of the scores in control group were 0, while they drop to around 40% in treated animals. The scores used for mild fibrosis (scores 1, 2 and 3) represent 23 to 27% in control animals, while they range from 37 to 43% in treated animals. Conversely, scores < 3, signaling moderate to strong fibrosis, are used more often (28 and 15% for females and males, respectively) for treated animals, while they represent a residual 4 and 2% for the control animals. These results can be visualized in Fig. 1.

Control animals

In females one animal presented an atypical renal mesenchymal tumour metastasis in the lungs, influencing the level of inflammation and fibrosis, and was therefore considered an outlier and excluded from the analysis. This resulted in 39 lung samples from female untreated animals, for a total of 1560 fields analysed. The mean Ashcroft score per animal was always < 1, while the median per animal was 0 in all cases but one (0.5). The mean per group of the mean Ashcroft score per animal was 0.60 (S.D. 0.20) and the mean value of the median was 0.01 (SD 0.08) (see Table 4 and supplementary material).

In males, again one animal was considered an outlier due to his fibro-histiocytic sarcoma of the lungs and not included in the analysis; furthermore, three more observations from another animal were not available due to insufficient sampling material, leading to a total of 1557 observed fields. The mean Ashcroft score per animal was < 1 in all cases but one (1.225), similarly the median per animal was 0 in all cases but one (0.5). The mean per group of the mean Ashcroft score per animal was 0.47 (SD 0.24) and the mean value of the median was 0.01 (SD 0.08) (see Table 4 and supplementary material).

In female and male controls, fibrosis was limited to a few perivascular, peribronchial or subpleural areas. Sporadic occurrence of minimal fibrotic lesions was compatible with the age of rats. The difference in the average Ashcroft score between the groups, 0.60 in females and 0.47 in males, could be attributable to the mean age at death (94.1 weeks for males and 101.9 weeks for females) (see supplementary material).

Fibrous FE-treated SD rats

Forty fields per animal were analysed and scored using an Ashcroft scale for a total of 1600 scores for treated males and females.

In female FE-treated group we observed a higher variability in the mean Ashcroft score per animal: 19 had mild fibrosis with an average score between 1 and 2 (47.5%); 17 had moderate fibrosis with an average score between 2.1 and 3 (42.5%); finally, 4 cases, including 3 rats with diagnosis of mesothelioma, had severe fibrosis with an average score > 3 (10%). In animals bearing mesothelioma, 2 cases had a mean score between 1 and 2; 2 cases > 2; and 3 cases > 4. In animals showing pleural plaques, 2 cases had a mean score > 2 and 1 case had a score between 1 and 2. Similarly, the median scores per animal were more heterogeneous, ranging from 0 to 7.5 (see supplementary material). The mean per group of the mean Ashcroft score per animal was 2.28 (SD 0.95), three times higher than control, and mean value of the median was 1.95 (SD 1.59).

In male FE-treated group, we observed a higher variability in the mean Ashcroft score per animal: 35 had mild fibrosis with an average score between 1 and 2 (87.5%), 3 had moderate fibrosis with an average score between 2.1 and 3 (7.5%) and 2 had severe fibrosis with an average score > 3 (5%). In animals bearing mesothelioma, 4 had an average score between 1 and 2, and 2 had an average score > 3. In animals showing pleural plaques, mean Ashcroft score did not exceed value of 1. Similarly, the median scores per animal were more heterogeneous, ranging from 0 to 7(see supplementary material). The mean per group of the mean Ashcroft score per animal was 1.58 (SD 0.87), three times higher than control, and mean value of the median was 1.05 (SD 1.22).

The statistical analysis of the group mean of the mean and median Ashcroft scores showed a statistically significant increase of fibrosis (p < 0.0001) in both male and female FE-treated animals compared to controls (Table 4).

In 24 out of 40 FE-treated female rats (60%) and in 27 out of 40 FE-treated male rats (67.5%), fibrosis was diagnosed at pleural and subpleural level. In animals with the highest score, fibrosis was also present into the alveolar parenchyma and not only in peribronchial or perivascular areas.

Fibrosis induced by FE-treatment was frequently associated with inflammatory reaction characterized by macrophages, granulocytes, and multinucleated giant cells, mainly found in the pleura. Mesothelioma histotypes differed in the degree of fibrosis: the sarcomatous histotype had very high score due to the high presence of collagen fibres compared to biphasic and epithelioid ones. For this reason, not all mesotheliomas achieved similar scores. Other fibrotic lesions observed in FE-treated groups were fibrous plaques, nodules, and pleural fibrosis. In both mesotheliomas and pleural plaques, the extension of fibrosis frequently affected the diaphragm.

Immunohistochemical (IHC) characterization of lung fibrosis

Alpha-SMA, COL1-A and vimentin, showed variable grades of positivity in mesotheliomas, while no marker was expressed in untreated controls. Vimentin was the most expressed marker, showing positivity also in mild lesions. Alpha-SMA was found to be the least expressed marker overall, but still positive in mesotheliomas and pleural plaques. Collagen type I was generally more expressed than alpha-SMA showing strong positivity in severe lesions and weak positivity in pleural fibrosis. The overall results underlined that mesotheliomas were characterized by high degrees of fibrosis. In fibrotic plaque or nodule and in fibrosis with fibre deposits, the positive staining for vimentin underlined mesenchymal cell activation and collagen production (Fig. 1). In mild lesions, a variable grade of positivity was observed only for vimentin, showing that the first stage of fibrosis development is the activation of mesenchymal cells.

Perls’ staining for the identification of iron deposits in lungs

Perls’ staining was performed to evaluate the presence of FE fibres and their chemical composition by the indirect staining of iron deposits in lung tissue samples and to understand the mechanisms of interaction between fibres and tissues.

In 27 out of 39 stained lung samples (14 females and 13 males, 69.2%), the presence of ferruginous bodies near or engulfing FE fibres was observed (Fig. 2). Of the remaining 12 cases, fibres were not present in 6 cases (3 females and 3 males, 15.4%). In the remaining 6 cases (3 females and 3 males, 15.4%), fibres were detected but ferruginous bodies were not found around them. Most cases with fibre deposits showed a concomitant formation of ferruginous bodies of different shapes and sizes covering either the entire fibre or part of it. In mesotheliomas iron deposits were observed in intercellular or intracellular spaces (Fig. 3).

Fig. 2

Case No. 101 Pleural fibrosis with deposits of fibers. A localized fibrosis was observed within a pleural plaque, in which deposits of fibrous FE are visible. Fibrosis is stained in green by Masson's trichrome (A). The IHC markers showed different expression: Alpha-SMA is poorly expressed and localized around the vessels (B); Vimentin is strongly expressed (C); COL-1A is weakly expressed in the area containing fiber deposits (D) (Magnification 40X)

Fig. 3

Case No. 125 (F) Fibrosis with fiber deposits. This image shows the formation of deep blue ferruginous/asbestiform bodies around the fibers. These formations have different shapes: segmented, rounded or needle-like. (PERLS stain, Magnification 40X)

Immunohistochemical and immunofluorescence analysis of ferritin in lungs

IHC analysis of ferritin and Perls’ staining showed comparable results. In 27 cases (69.2%), ferruginous bodies showed positivity for both ferritin and Perls’ staining. In 6 cases (15.4%), fibres were not observed, and both the staining were negative. In the remaining 6 cases (15.4%), fibres were observed, but no ferruginous bodies or iron deposits were found. Fibrous plaques and fibrosis with fibre deposits showed strong positivity for ferritin around fibres and in extracellular matrix. In mesotheliomas, ferritin expression was lower than in non-neoplastic lesions. In spindle cell mesothelioma, ferritin deposits were observed within the tumour and among the neoplastic cells. In fibrosis, in which no fibres or iron deposits were found, the expression of ferritin was absent. Analysis of ferritin expression was also performed by IF, with ferritin highlighted in red while nuclei in blue.