Malignant Mesothelioma In Situ: A Controversial Diagnostic Entity – A Review

Abstract

Malignant mesothelioma is a rare aggressive malignancy of serosal surfaces that is strongly associated with exposure to asbestos. The pathological diagnosis of malignant mesothelioma can be difficult to distinguish from reactive benign hyperplasia. Mutations in BAP-1 and CDKN2a distinguish mesothelioma from reactive hyperplasia. An in situ growth phase of mesothelioma until recently was difficult to ascertain due to limits of histological assessment and because mesothelioma tends to spread diffusely along serosal surfaces making sampling for invasion impossible without extensive resection. The current WHO classification of thoracic tumors recognizes mesothelioma in situ as a distinct entity based on histological, genetic, and clinical features. This chapter reviews the topic and concludes that the diagnosis of malignant mesothelioma in situ should be limited to patients eligible for radical resection to confirm the putative diagnosis.

1. Introduction

Malignant tumors result from dysregulated clonal cell proliferations. Malignant tumors can potentially invade local tissues and metastasize to distant sites. There are many pathways leading to malignant transformation and the spread of tumor. Most solid epithelial tumors pass through a stage of in situ neoplasia, in which disordered cellular proliferation occurs locally, and in the absence of invasion of underlying tissues. Pathologists are trained to recognize in situ malignant neoplasia based on its histological features, including non-disruption of the cell basement membrane. The clinical importance of distinguishing in situ malignancy is that it may be cured by surgical resection in a high percentage of cases.

Malignant mesothelioma is a rare malignancy of serosal-lined surfaces [1]. Most mesotheliomas occur in the pleura (85%), with the remainder in peritoneum (~10%), pericardium and tunica vaginalis (each less than 2.5%). Mesothelioma was rarely described in the medical literature prior to 1960, when Wagner described a large cohort of cases in South African asbestos miners and those living in proximity to the mines. Subsequently, a large percentage of mesotheliomas have been demonstrated to be causally linked to prior exposure to asbestos, the latter a naturally occurring fibrous magnesium silicate that was used extensively as an insulating material in the construction and other trades in the twentieth century. There is a prolonged latency for the development of malignant mesothelioma, on average greater than 30 years following first exposures to asbestos. Other recognized causes of mesothelioma are rare and include therapeutic radiation and the mineral erionite that was used in the construction of homes in the Cappadocia region of Turkey [2].

Mesothelial cells line normal serosal surfaces. They secrete acid mucins and regulate cavitary fluid accumulation of lubricant fluid. In response to inflammation, normal mesothelial cells proliferate and show enhanced fluid secretion, leading to exudative effusions that bring patients to clinical attention. Reactive mesothelial proliferation can be difficult to differentiate histologically from mesothelial malignancy. Mesothelial cells are pluripotential cells; and under neoplastic conditions, they may give rise to tumors with epithelioid, sarcomatoid, or mixed (biphasic) histologies. Until recently, unequivocal evidence of soft tissue invasion was required to distinguish malignancy from benign reactive mesothelial proliferations. Unfortunately, the propensity of malignant mesothelioma to spread diffusely along serosal surfaces and to invade adjacent structures makes curative surgical excision virtually impossible. Consequently, nearly all patients with malignant mesothelioma die from their disease.

2. Genetics of mesothelial malignancy

Malignant mesothelioma are heterogeneous with complex genetic, chromosomal, and epigenetic changes. Mesotheliomas are often polyclonal neoplasms, likely reflecting a “field effect” induced by asbestos. They also display relatively low mutation burdens, compared to most adult solid tumors [3].

The Cancer Genome Atlas program studied 74 mesotheliomas for genetic alterations using next-generation sequencing (NGS), whole-exome sequencing (WES), messenger RNA expression, methylation analysis, microRNA expression, exomes, reverse-phase protein array, and transcription factor analyses. They observed frequent characteristic mutations in BAP-1, CDKN2A, NF2, TP53, LATS2, and SETD2 [4]. Bueno et al. using NGS [5] and by Hmeljak et al. by WES [6] confirmed this mutation profile, suggesting that the development of mesothelioma may be due to mutations in a limited number of genes [1]. In this regard, Badhai et al. demonstrated that concomitant deletions of BAP1, NF2, and CDKN2A genes in mice resulted in malignant mesothelioma in all animals, and that the knockout of both CDKN2a and NF2 genes yielded mesothelioma in 75% [7]. However, knockout of each of these three genes individually failed to yield malignancy. Currently, there is no evidence that neoplastic mutations capable of causing mesothelioma occur spontaneously in nature in the absence of an additional environmental stimulus, e.g. asbestos. In this regard, Yoshikawa et al. found that chromothripsis, i.e. chromosome shattering followed by random chromosomal rearrangement, may be the cause of multiple genetic alterations observed in mesotheliomas caused by asbestos [8].

Roughly, half of all cases of malignant mesothelioma show biallelic mutations in BAP-1 (BRACA-associated protein-1), and 60–80% of mesothelial malignancies show biallelic deletions in the CDKN2A gene. BAP-1 codes for a ubiquitin carboxy-terminal hydrolase and is a member of the deubiquitylase family of proteins. The gene maps primarily to the 3p21.3 chromosome and its encoded protein can be localized to both the nucleus and cytoplasm of mesothelial cells. BAP-1 acts as a tumor cell suppressor via several pathways, and its functional loss drives tumor cell proliferation [5]. Families showing characteristic autosomal BAP-1 germline mutations show an increased incidence of a variety of malignancies, most characteristically uveal melanoma, malignant mesothelioma, and renal cell carcinoma [1]. However, there is no compelling evidence to suggest that germline mutations alone, in the absence of environmental carcinogen, can give rise spontaneously to mesothelioma.

Biallelic acquired deletions of CDKN2A (cyclin-dependent kinase inhibitor 2A) are associated with mesothelioma. Suppression of CDKN2A transcription is a common feature of a variety of tumors and is currently the most common genetic defect observed in malignant mesothelioma [9].

3. Invasion and metastasis

Malignant transformation can occur without tumor invasion and metastasis and may precede it in many cases. The steps leading to tumor invasion and spread are complex and include both genetic and epigenetic changes. The mammalian organism is divided into tissue compartments separated by extracellular matrix, i.e. basement membrane, and interstitial stroma [10]. Basal epithelial cells normally attach to basement membrane, or as in the case of mesothelial cells, directly to subjacent interstitial matrix. In normal tissues, these cell populations do not intermix.

On the “other” side of the basement membrane, the interstitial stroma includes an array of cells, including fibroblasts, myofibroblasts, and their secreted matrix. Like epithelium, nerve cells, muscle cells, and blood vessels are surrounded by a continuous extracellular matrix. During tissue remodeling, benign proliferative disorders, and carcinoma in situ, the cell populations on either side of this cell/matrix boundary do not intermix. But in malignant invasive tumors, a neoplastic cell, or group of cells, will penetrate the local stroma. It can then enter lymphatic or vascular circulations and subsequently migrate to distant sites, extravasate, and then proliferate as a secondary colony. Invasive tumor cells disobey the normal order of tissue/matrix boundaries and proliferate where they normally do not belong.

Tumor invasion by malignant cells is an active process [10]. Interactions of the tumor cell with basement membrane and interstitial matrix require initial attachment, the dissolution of matrix, and cellular migration. In this regard, tumor cells secrete lytic enzymes or induce the host to elaborate proteinases that degrade matrix adhesion molecules. Locomotion propels the tumor cell across the basement membrane and stroma through a zone of proteolyzed matrix. Directional movement is regulated by cell surface ligand binding and mobilization of cytoskeletal elements that interact with the cell membrane surface. Tumor cell motility is modulated by cytokines, and direction of movement is influenced by local chemoattractants.

Little is known about the specific pathways that promote tissue invasion and metastasis in malignant mesothelioma. Malignant mesothelioma cells produce collagens, and the prognosis of mesothelioma may be related to the expression of matrix metalloproteinases. Malignant mesothelioma tumors also induce immune responses from their hosts. Chronic inflammation is manifested by the presence of tumor inflammatory cells and the local release of cytokines. Mesothelioma cell growth also requires neoangiogenesis to provide nutrients to proliferating cells and supportive matrix. Growth signals and loss of tumor-suppressor genes may provide the growth advantage that leads to tumor cell proliferation.

4. Distinguishing malignant mesothelioma in situ

The existence of an in situ phase of malignant mesothelioma has long been postulated but difficult to prove. But as a consequence of tumor genetic analysis, it is now possible to identify mesothelial cells that harbor characteristic genetic mutations that are only seen in malignant cells and not seen in “atypical” but benign mesothelial proliferations. These include loss of BAP-1 expression and CDKN2A (p16) deletions. The loss of BAP-1 is most commonly present as an acquired somatic mutation (~50% of epithelioid mesotheliomas) but can occur rarely as a germ-line defect within families [11]. Distinguishing benign from malignant mesothelial cells via genetic analysis in cytology and tissue biopsy specimens currently allows distinguishing mesothelioma in situ from benign mesothelial proliferations. Recent studies suggest that immunostaining for MTAP (metalloadenosylphosphorylase) is a reliable substitute marker for loss of expression of CDKN2A/p16 by FISH, and absence of MTAP expression by immunostaining in cytology and biopsy specimens strongly suggests malignant transformation [10]. In a similar vein, loss of nuclear BAP-1 protein expression by immunostaining correlates with loss of BAP-1 gene and is a reliable marker of neoplastic transformation (see Figure 1).

Figure 1.

A 74-year-old man presented with recurrent right pleural effusions. Cytological analysis of the fluid showed (A) clusters of atypical mesothelial cells with loss of BAP-1 nuclear expression (arrow) by immunohistochemistry consistent with malignancy. Retention of BAP-1 immunostaining (brown nuclear pigment) is seen exudate inflammatory cells. Pleurectomy/decortication revealed (B) a single layer of mesothelial cells with loss of BAP-1 expression (not shown) lining the pleural surfaces with no evidence of invasion. The patient has been clinically free of tumor for 18 months.

In 2018, Churg et al. described 10 patients with mesothelioma in situ, which they defined as (1) a single layer of surface mesothelial cells (2) showing loss of BAP1 nuclear immunostaining, with (3) no evidence of tumor invasion by radiographic imaging for at least 1 year [12]. Nine of their cases were pleural mesotheliomas, with one peritoneal mesothelioma. Most patients were biopsied for repeated pleural effusions of uncertain etiology. In two patients, mesothelioma in situ was an incidental finding in lungs that had been resected for pulmonary carcinoma. CDKN2A showed deletions by FISH analysis in only one of eight cases. Invasive malignant mesothelioma developed in seven patients at 12 to 92 months following the diagnosis of in situ mesothelial neoplasia, and there was no clinical evidence of invasive mesothelioma in three patients at 12, 57, and 120 months of follow-up. The authors concluded that mesothelioma in situ has a high risk of becoming invasive over time, but that curable interventions might in theory be possible in such cases.

Elliott et al. [13] reported a case of malignant peritoneal mesothelioma that had apparently progressed from mesothelioma in situ over a 10-month period in a 24-year-old woman with advanced endometriosis. Initial surgery showed deeply infiltrative endometriosis with progestin effect. Postoperatively, the patient had intractable pelvic pain and vaginal discharge. Imaging studies were negative, but laparoscopy 10 months later revealed vesicular lesions on the omentum and pinpoint white lesions studding the small bowel, appendix, and pelvic peritoneum. A diagnosis of epithelioid mesothelioma was established based on biopsy of the omentum and was confirmed by immunohistochemical loss of BAP1 expression. Upon genetic testing, the patient was found to have a germline mutation in BAP1.

5. The WHO classification of mesothelioma in situ

The most recent WHO classification of thoracic malignancies includes malignant mesothelioma in situ as a distinct diagnostic entity. According to the new criteria “mesothelioma is clinically suspected in patients presenting with non-resolving pleural effusions in the setting of heavy asbestos exposure with or without pleural plaques and in patient with familial predisposition. The diagnosis is based on a combination of clinical imaging and pathological features.” It should be noted that this differs from the standard criteria for most in situ malignancies, which are ascertained exclusively by the absence of pathological evidence of tissue invasion. According to the WHO, the diagnosis of mesothelioma in situ requires “extensive” thoracoscopic sampling of tumor, and small biopsies or cytological sampling are judged insufficient to establish this diagnosis.

Although the diagnosis of mesothelioma in situ cannot be made by histological examination alone, any histological evidence of invasion excludes the diagnosis. Loss of BAP-1 expression immunohistochemically (or alternatively loss of CDKN2A either by FISH or loss of MTAP expression by immunostaining must be present) as loss of these genetic markers reliably distinguish mesothelial malignancy from benign hyperplasia.

To date, only a handful of cases of malignant mesothelioma in situ have been reported. These have been predominantly in the pleura and in men in the seventh decade. Apparently, demographics will not distinguish in situ mesothelioma from invasive malignancy. Klebe et al. [14] noted that an international survey of 34 pulmonary pathologists, with an interest in malignant mesothelioma diagnosis, exhibited marked inconsistencies in establishing the diagnosis of in situ mesothelioma, despite the published WHO guidelines.

Serious problems apparently persist with respect to the confidence in which this diagnosis is made, as well as in with how to approach this “entity” therapeutically. It may be premature to accept mesothelioma in situ as a discrete entity in (1) the absence of clearly defined criteria with respect to what constitutes adequate sampling and (2) the lack of consensus as to how best to approach it therapeutically. The current WHO criteria apply to mesotheliomas of the pleura, presumably because it may be too difficult to establish this diagnosis with confidence in the larger peritoneal cavity.

Accurate sampling to exclude microscopic foci of invasion is virtually impossible to achieve in the absence of complete tumor excision with detailed pathological examination. The WHO criteria do not include validated criteria for how extensively to sample serosal surfaces, although it recognizes that limited sampling will lead to overdiagnosis of this entity.

Although having molecular markers of mesothelial malignancy is a substantial scientific advance, enthusiasm should be balanced by consideration of the clinical implications of the diagnosis of mesothelioma in situ. As previously noted, for most localized solid epithelial tumors, diagnosing an in situ carcinoma will trigger surgical excision with high cure rates. The question is whether malignant mesothelioma in situ is equally amenable to such an approach. Unfortunately, many patients with malignant mesothelioma, both invasive and in situ, are elderly with significant co-morbidities so that the extensive surgical resections necessary to extirpate all areas of in situ mesothelioma may be precluded. Considering the high rate of developing invasive disease within five years, the clinical value of making a diagnosis of mesothelioma in situ is questionable, unless the patient is a candidate for radical surgery. When curative therapeutic intervention is not possible, the diagnosis of mesothelioma in situ has no demonstrated clinical implication.

For this reason, it is argued that a diagnosis of mesothelioma in situ should be reserved for cases for whom curative excision is feasible. Only then the diagnosis can be pathologically confirmed and the potential value of the diagnosis confirmed by long-term tumor-free survival. Despite understandable enthusiasm for designating a new category of mesothelial neoplasia based primarily on genetic evidence, cases meeting the current criteria who are not surgical candidates might best be diagnosed as “malignant mesothelial neoplasia,” with the expectation that their tumor will likely become frankly invasive at some future point in time in the absence of curative treatment. Patients meeting the current WHO criteria for mesothelioma in situ but who are not surgical candidates should be assigned to experimental chemotherapy or immunotherapy protocols to determine whether sustained responses can possibly be achieved. Unfortunately, intercurrent mortality from other causes may complicate the evaluation of this “entity.”

6. Conclusion

Malignant mesothelioma is a rare and deadly malignancy. Five-year survival rates are negligible, even following surgical and chemo/immunotherapeutic interventions. For this reason, the possibility of defining a preinvasive stage of malignant mesothelioma theoretically offers the possibility of therapeutic cure. However, the diffuse nature of the disease and the advanced age of most patients with mesothelioma in situ represent obstacles to accurate diagnosis and curative treatment. Patients who meet current WHO criteria for this diagnosis should be carefully selected and treated on protocols to determine whether surgery and/or chemo/immunotherapies can yield long-term survival. Identifying confirmed cases of in situ mesothelioma may also allow researchers to determine molecular/genetic pathways responsible for the invasive behaviors of malignant mesothelioma and potentially provide new molecular targets for therapeutic intervention.

Acknowledgments

The author would like to thank Dr. Siobhan Cashman for her assistance in the preparation of the manuscript.

Conflict of interest

Dr. Kradin has testified as an expert witness in cases of mesothelioma attributed to asbestos.