**5. Fb-3 as biomarker in asbestos-related pathologies**

One of the earliest studies involving the role of Fb-3 in cancer was carried out in 2009 in the United States on gliomas [19]. Following some preliminary investigations, the authors had hypothesized that gliomas' local invasiveness could be caused by an interaction between mesenchymal proteins and some specific neural matrix proteins. In fact, unlike other central neuro system neoplasms (CNS), characterized by an expansive growth with shifting and compression of the surrounding parenchyma, gliomas show an infiltration type of growth. The extracellular matrix in the CNS normally contains high quantities of ialuronic acid and negatively charged proteoglycans, but low quantities of fibrillar proteins, which may support cellular adherence and mobility. In an attempt to identify the humoral signs, which could contribute to gliomas' peculiar invasiveness (probably due to an altered relationship between cells and extracellular matrix), researchers have stimulated tumoral cells *in-vitro,* combining mesenchymal elements (fibronectine) with others, specific of the neural matrix (brevican) and, by using a micro-array, examined which genes came out overexpressed. Results showed a remarkable rise of Fb-3 expression in the glioma.

Successive studies about the role of Fb-3 showed mixed results: some observed an antagonist effect toward tumoral angiogenesis (reducing so its aggressiveness); others, as in the case of pancreatic adenocarcinomas and gliomas, observed Fb-3 expression rise associated with an increased vascular growth factor (VEGF) and tumoral growth. **Table 1** reports some studies on Fb-3 and of different kind of cancers.

A research of 2011 on colorectal cancer detected an Fb-3 *downregulation* in the cancerous tissues compared with the adjacent healthy ones. Furthermore, the Fb-3 *downregulation* negatively correlated with the prognosis, tumor stage, lympho-node metastatis, and reduced time gaps free from the illness.

In a following study, Fb-3 plasma levels were tested in colon-cancer subjects, comparing them with a control group made up of healthy subjects. The Fb-3 resulted significantly reduced in tumor-stricken subjects, in a directly proportional manner with lymphonode metastases and, at length, the tumoral mass and in general with the neoplasm stage.

In intestinal tumors, Fb-3 *downregulation* seems then to show a worsened prognosis due to a reduced anti-angiogenic action.

In the light of what has been said, Fb-3 is thought to play, depending on the tumor type, a pro-angiogenic role (gliomas, cervix carcinoma) or an anti-angiogenic one (colon carcinoma). This apparent paradox may derive from a different behavior of this glycoprotein in relation to factors such as tissue histological characteristics and tumor micro-environment. The Fb-3 bond with TIMP-3 might interfere with that between VEGF to its type 2 receptor (VEGFR-2), causing the inhibition of tumoral angiogenesis. Besides, it has been observed that Fb-3 is able to competitively link the EGF receptor (EGFR) compared with EGF, so activating intra-cellular pathways (MAPK, Akt) in pancreatic adenocarcinoma. Finally, a few studies have concluded that the


*Fibulin-3 as a Biomarker of Pleuric Involvement by Exposure to Fibers DOI: http://dx.doi.org/10.5772/intechopen.104448*

#### **Table 1.**

*Studies exploring Fb-3 in relation to cancer type.*

EFEMP1 gene activation, due to the hypermethylation of its promoter, also occurs in several cancer types (lung, prostatic, colorectal, nasopharyngeal, and hepatocellular).

These data seem to point out that a reduced expression of this gene may be involved in carcinogenic and/or tumor growth processes. It seems however evident how the exact role of Fb-3 in tumoral growth still remains to be clarified and needs further research.

Recent studies on the pathophysiological role of Fb-3 in malignant mesothelioma (MM) are also taking into account this glycoprotein as a possible marker for early diagnosis and/or pathology follow-up.

This research falls within a larger assessment context of potential biomarkers in MM early diagnosis.

MM is a fatal tumor, with a long latency and aspecific symptoms, which often end up in a late diagnosis. MM is causally correlated with exposure to asbestos or asbestiform fibers. MM cases worldwide are definitely increasing. In Italy, an incidence peak is expected within 2025 [31, 32].

Actually 25% of MM is caused by professional exposure, 25% through indirect exposure of family members, and 50% from exposure to fibers in the surrounding environment [33].

MM patients survive averagely 6–18 months from diagnosis. However, it has been noticed that, if an early diagnosis is made, survival may even go beyond 5 years. Unfortunately, today there are still no effective prevention systems and screening procedures for this pathology [31–34].

Periodical X-ray exams have always been hard to do, due to: long latency of the disease (14–45 years); limited resolution of present techniques, especially for lesions at an early stage; exposure to ionizing radiations (justification principle). It is then clear how finding humoral biomarkers with high sensitivity and specificity might significantly enhance the prognosis of this disease.

Scientific debate on the eligible molecules has been going on for long, with no definite results. Indeed, none of the biomarkers studied seems to meet the requirements needed [35, 36].

The first study to propose using Fb-3 as a possible MM biomarker was conducted by Pass et al. [3]. The intent of the study was to analyze Fb-3 reliability compared with mesothelin, a protein already thoroughly studied as a biomarker, which had however shown no adequate sensitivity (47%) in recognizing MM cases. Plasma and effusion samples from patients with pleural MM, plasma samples from persons who had been exposed to asbestos but did not have MM, and plasma and effusion samples from patients with pleural effusions not due to MM were analyzed.

In this study performed on MM patients, sampling was carried out in the United States, at the Wayne State University, from 1998 to 2005, and at New York University Langone Medical Center, from 2005 to 2011, the "Detroit Cohort" and the "New York Cohort," respectively.

The study also assessed patients with other neoplasms, in order to improve Fb-3 specificity. Altogether, 20 ovarian cancer, 20 glioblastoma, and 31 prostatic carcinoma patients were evaluated. Furthermore, 43 healthy subjects were used as control group (selection criteria included absence of previous exposure to asbestos and other neoplastic pathologies).

In conclusion, plasma Fb-3 levels can distinguish healthy persons with exposure to asbestos from patients with MM. In conjunction with effusion Fb-3 levels, plasma Fb-3 levels can further differentiate MM effusions from other malignant and benign effusions [3].

On the whole, there were 11 studies dealing with concentrations of Fb-3 in human beings, and they were performed on: MM tumoral tissue; pleural exudate; serum and plasma (see **Table 2**).

The surveys on MM patients' tumoral tissues were conducted by Pass et al. [3] and Caltabiano et al. [6].


*Fibulin-3 as a Biomarker of Pleuric Involvement by Exposure to Fibers DOI: http://dx.doi.org/10.5772/intechopen.104448*

#### **Table 2.**

*Studies exploring Fb-3 in pleural fluids and peripheral blood.*

In Pass et al. [3], the immunohistochemical analysis enabled to give a score to the nuclear as well as to the citoplasmatic positivity, taking into account both the number of positive cells and the positivity intensity. The authors detected Fb-3 nuclear and citoplasmatic expression in 100% of MM samples (26/26); the scores for intensity were similar both for the epithelial subtype, and the sarcomatoid and the sarcomatoid/epithelial mixed variant ones.

Comparison among the epithelial-histological, the epithelial-biphasic-histological and the sarcomatoid subtypes showed similar scores as far as the coloring intensity was concerned (mean score 7.7 ± 0.6 and 6.9 ± 0.8, respectively P = 0.87); and so did it with purely sarcomatoid- histological subtypes (6.6 ± 1.1; P = 0.62). The total coloring score (nuclear and citoplasmatic) turned out constantly higher in MM samples than in those detected in other pleural neoplastic forms (7.4 ± 0.5 vs. 2.4 ± 0.8; P < 0.001).

In Caltabiano et al. [6], Fb-3 immunohistochemical expression was assessed on tumoral tissues of six MM patients, previously exposed to fluoroedenite (FE); a natural, asbestiform fiber discovered in lava rock stone used as construction material in Biancavilla's municipality, on the slopes of Mount Etna.

Outcomes showed immunoexpression similar in the epithelial histological subtypes (three cases) and in the epithelial biphasic histological and sarcomatoid subtype (three cases) (see **Figure 4**).

The analysis of Fb-3 concentration in the pleural exudate was carried out in four studies: Pass et al. [3]; Creaney et al., [12]; Agha et al. [37]; Battolla et al. [8].

#### **Figure 4.**

*IHC expression of Fb-3. (A) Hematoxylin and eosin (H&E) stained section of an epithelioid mesothelioma composed of glandular structures exhibiting diffuse and strong staining for Fb-3 (B). Immunohistochemical staining for Fb-3 in the same case depicted in (A): Neoplastic cells within glandular structures exhibit diffuse and strong nuclear and cytoplasmic staining for Fb-3. (C) H&E stained section of another case of epithelioid mesothelioma composed of closely packed glands showing diffuse and moderate staining for Fb-3 (D). Immunohistochemical staining for Fb-3 in the same case depicted in (B): Neoplastic cells show a diffuse staining for Fb-3 of moderate intensity. (E) H&E section of a biphasic mesothelioma composed predominantly of neoplastic spindle-shaped cells; (F) this case exhibited a diffuse staining for Fb-3; however, the staining intensity was recorded as weak.*

Pass et al. [3] observed significantly higher concentrations of Fb-3 in MM subjects' pleural exudate than those detected in benign exudates or derived from other neoplasms. Therefore, the Fb-3 concentration allowed to tell MM subjects from all the others (area below the curve-AUC = 0.93), both in benign (AUC = 0.93) and in malignant pathologies (AUC = 0.94). Moreover, Fb-3 levels did not significantly differ between those patients (n = 22) who had received presurgery chemotherapy and those (52) who had not (617.4 ± 72.5 vs*.* 703.6 ± 42.6 ng/ml).

Fb-3 significantly correlated with the progress of the disease and made it possible to distinguish those patients (n = 54) who underwent citoreductive surgery in stage I-II (n = 21), from those (n = 33) with III-IV stage disease (576 ± 67 ng/ml vs*.* 765 ± 55 ng/ml, P = 0.04).

An Fb-3 = 733.4 ng/ml cutoff, measured at the time of surgery in all subjects (n = 69), correlated in an inverse proportional way with patients' survival.

Creaney et al. [12] detected Fb-3 values and mesotheline in the pleural exudate of 153 patients: 82 had MM, 36 had pleural exudate caused by other neoplastic pathologies; 35 had benign exudates.

The MM patients' samples were collected within a month from diagnosis, prior to any kind of treatment. Fb-3 levels ranged between 17 and 5748 ng/ml. No significant difference was detected in Fb-3 levels among the three groups under exam; in detail, 63% of benign exudate samples exceeded the 346 ng/ml cutoff. Statistical analysis showed no difference in Fb-3 levels according to the pleural liquid protein composition (exudate or drained) and/or with blood.

Exudates coming from MM patients with biphasic or sarcomatoid histology showed significantly higher levels of Fb-3 (1331, range 538–2486 ng/ml) than the epithelial ones (426, range 171–1709 ng/ml; P = 0.018), also in those patients who had

#### *Fibulin-3 as a Biomarker of Pleuric Involvement by Exposure to Fibers DOI: http://dx.doi.org/10.5772/intechopen.104448*

had a citological-based diagnosis (298, range 155–881 ng/ml; P = 0.002). No significant difference was observed in Fb-3 levels according to the stage of the disease. Altogether, the Fb-3 study results in the pleural exudate showed a 59% sensitivity and a 52% specificity, considering 346 ng/ml. as threshold. The 0.588 AUC enabled to tell MM patients from all the others.

As regards mesothelin levels, they were remarkably higher in MM patients than in those with benign exudates (P < 0.001) and in others with exudates caused by other neoplasms (P < 0,001).

Creaney and colleagues concluded that mesothelin gave out a 58% sensitivity and a 96% specificity, as well as a better diagnostic accuracy, compared with Fb-3 in pleural exudates of MM patients.

A study carried out by Agha et al. [37] analyzed 45 patients with pleural exudate, of whom: 25 MM cases, 11 secondary pleural metastases (3 cases of not-small-cell lung cancer, 2 breast cancers, 3 colon cancers, 1 case of kidney cancer, and 2 cases of limphoma), and 9 patients with benign origin pleural exudates (5 tubercolosis, 1 pneumonia, and 3 pleurisies). MM patients showed significantly higher Fb-3 levels (331 ± 32.64 ng/ml) than those with pleural exudate derived from secondary metastases (153.01 ± 60.32 ng/ml). The difference between these parameters turned out to be statistically significant (P < 0.001).

The results highlighted that with a 150 ng/ml cut-off (AUC = 0.878; 72.3% sensitivity, 80% specificity), it was possible to tell MM patients from those with pleural metastatic pathology.

Besides, exploiting a 127.5 ng/ml cut-off (AUC = 0.909; sensitivity 88%, specificity 77.8%) it was possible to distinguish MM from the pleural benign exudate.

In a recent study, Battolla et al. compared Fb-3 and mesothelin levels in MM patients' pleural liquid with that obtained from patients with pleural pathologies, both benign and malignant, other than MM. 120 subjects underwent thoracenthesis between 2008 and 2011. Among these, 33 had MM, 64 had benign pleura lesions and 23 secondary pleural metastases. Fb-3 and mesothelin concentrations were assessed in ELISA. Results showed Fb-3 levels substantially similar in all subjects (P = 0.174), whereas mesothelin levels were significantly higher in MM subjects than others (P = 0.001).

The analysis of Fb-3 concentration in peripheral blood was conducted in seven surveys on plasma and three on serum.

In Pass et al.'s study (2012), Fb-3 plasma values were assessed. The study sample included: 92 MM patients; 136 exposed to asbestos with no cancer; 93 patients with nonrelated MM pleural exudate; 43 healthy subjects as control group. The study was carried out in two separate cohorts: "Detroit cohort" and "New York cohort."

Outcomes highlighted that Fb-3 average plasma levels enabled to significantly distinguish asbestos-exposed subjects from those with nonrelated MM exudate and from MM ones, in both cohorts. Fb-3 concentrations in MM "Detroit cohort" patients were similar to those of the "New York cohort" (105.0 ± 7.1 vs*.* 112.9 ± 7.6 ng/ml; P = 0.63). Fb-3 plasma levels did not significantly differ between the 44 MM patients, who had had presurgery chemotherapy and the 48 who had not (117.9 ± 8.1 vs*.* 101.1 ± 6.9 ng/ ml; P = 0.12).

Fb-3 plasma level allowed to tell MM patients from those affected from other cancers or even those with pleural exudate (not MM-related), both benign and malignant. Finally, comparing the 28 patients at stage I-II of MM with the asbestosexposed ones, with AUC = 0.99 and cutoff = 46.0 ng/ml, a 100% sensitivity [95% IC, 87.7–100] and a 94.1% specificity [95% IC, from 88.7 to 97.4] were reached. Fb-3 levels of MM patients went down after surgery in 100% of cases (18 out of 18).

Contrary to expectations, the authors found poor correlation between Fb-3 levels found in plasma samples and those detected in the pleural exudate of each MM patient (n = 17) (P = 0.98), as well as in the plasma and pleural exudate of 15 patients who had not MM-related exudate (P = 0.27).

Among the conclusions, the authors suggested using plasma samples instead of serum ones so as to assess Fb-3 blood levels, since the presence of two potential trombine cleavage sites could compromise the validity of the exam. Despite the encouraging results obtained by Pass's survey, further experiments gave out mixed outcomes.

A cohort of 153 patients (of whom 82 having MM) was investigated by Creaney et al., reporting a 22% sensitivity and a 95% specificity for plasma Fb-3 (cutoff = 52 ng/ ml, AUC = 0.671). These values were definitely lower than those obtained, with the same patients for mesothelin (sensitivity 56%; specificity 95%—AUC = 0.816), which on the contrary seems to have a decisively better diagnostic accuracy on plasma samples. Although in this study mesothelin resulted superior to Fb-3 as to its diagnostic worth, the authors considered the latter superior from a prognostic point of view. Indeed, Fb-3 high levels correlated negatively with the patient's prognosis. A possible explanation of this might depend on an Fb-3 higher expression by biphasic and sarcomatoid histotypes, which are generally characterized by a worse prognosis. Instead, mesothelin is mainly expressed by the epithelial histotype, with a better prognosis.

An Egyptian study [37] conducted on a small cohort of 45 subjects reported a 100% sensitivity and a 78% specificity in differentiating MM cases (n = 25) from nonmalignant pleural pathologies (n = 9), and an 88% sensitivity and 82% specificity in distinguishing MM from other forms of pleural cancer (n = 11). It is necessary, though, to point out that the authors, when evaluating Fb-3, used a nonspecified test and internally agreed cutoffs.

Corradi et al. assessed the concentration of Fb-3 and other protein biomarkers in the serum of four groups of patients: subjects previously exposed to asbestos and suffering from asbestosis; patients with MM; patients with not-small-cell lung carcinoma (NSCLC) and a control group, which showed no evidence of neoplastic pathologies. The results highlighted higher levels of Fb-3 in MM patients than the NSCLC group (P < 0.01) and the control (P < 0.05). However, Fb-3 values in MM patients did not significantly differ from those of subjects with asbestosis. The small number of patients in the study is the main weakness of these results.

A prospective survey carried out by Kaya et al. [11] examined 43 MM patients (primary involvement: 39 pleural, 4 peritoneal mesothelioma) and 40 controls. Results showed Fb-3 serum levels equal to 90.3 ± 42.1 and 17.8 ± 12.7 ng/ml, respectively (P < 0.001). A 36.6 ng/ml cutoff indicated a 93% sensitivity and a 90% specificity.

Napolitano et al. [9] analyzed levels of *high mobility group box protein* 1 (HMGB1) and Fb-3 in blood samples of 22 MM subjects, 20 others with documented exposure to asbestos, 13 with benign pleural exudate, 25 with malignant exudate (other than MM) and 20 controls. The authors concluded that the combination of HMGB1 and Fb-3 provided higher sensitivity and specificity in differentiating MM patients from others with benign or malignant pleural pathologies.

MM etiology usually involves professional and/or environmental exposure to asbestos. In an attempt to spot any possible differences among the MM types derived after environmental exposure, compared to the more frequently documented professional one, Demir et al. recruited a cohort of MM patients (n = 42) derived after

*Fibulin-3 as a Biomarker of Pleuric Involvement by Exposure to Fibers DOI: http://dx.doi.org/10.5772/intechopen.104448*

#### **Figure 5.**

*Western blot analysis of Fb-3 protein level evaluated in primary human lung fibroblasts exposed to 10, 50 and 100 μg/ml of FE fibers for 72 h.*

environmental exposure to asbestos and compared them with two control groups: the former composed of healthy individuals (n = 48) who had no previous, documented exposure to asbestos, with a normal chest X-ray exam; the latter, (n = 48) composed of subjects with documented environmental exposure to asbestos for at least 15 years, who showed no X-ray documented pleural plaques. The authors detected significantly higher values of Fb-3 in MM patients' serum than in those who were just exposed to asbestos and the nonexposed control group.

Several investigations conducted by the working group (Caltabiano, Ledda and Loreto) directed by Rapisarda et al. [1, 6, 38, 39] analyzed the role of Fb-3 as biomarker in workers exposed to FE. Fb-3 plasma concentrations were measured in the blood of the FE-exposed workers and in a control group (non-exposed). In 52% of exposed subjects pleural plaques were detected. Fb-3 plasma concentrations resulted 12.96 e 5.29 ng/ml, respectively, in the exposed subjects compared to the control (P < 0.001).

The results highlighted a high predictive value of Fb-3 plasma levels in relation to the presence of pleural plaques.

Another survey revealed an Fb-3 increased expression in human mesothelial cells after exposure to FE. Moreover, the Fb-3 levels in the peripheral blood of 40 workers exposed to asbestos were analyzed and compared with those of professionally FE-exposed ones.

Also in this case, results showed Fb-3 higher levels in the FE-exposed group with pleural plaques than in those asbestos-exposed workers who did not show any pleural and/or parenchymal lesions.

At the same time, FE fibers were used to stimulate mesothelial cell cultures. Results showed an Fb-3 hyper-expression after exposure to FE even at low concentrations (see **Figure 5**).
