We received responses from 55 of the 68 surveyed institutions. Forty-five of the received surveys were returned by academic institutions, with community hospitals (7), military institutions (2), and reference laboratories (1) accounting for the remainder. The volume of “positive for malignant mesothelioma” and “suspicious for malignant mesothelioma” cases varied from fewer than 5 in the last 10 years (10% of respondents) to more than 50 (25% of respondents).
Of the respondents, 35 (63%) replied that they do make a definitive diagnosis of MM in effusion material. The vast majority of this group (94%) stated that they will do so in the context of both a primary diagnosis and recurrence. The most commonly cited feature useful in making a diagnosis was architectural atypia in the form of large groups of mesothelial cells with knobby borders. The next most commonly cited features were high cellularity in effusions composed of mesothelial cells and effusions containing mesothelial cells with marked cytologic atypia.
In cases where MM is suspected in a frankly malignant effusion, all respondents reported using a panel of IHC stains to rule out adenocarcinoma. Slightly more than half (55%) reported using ancillary studies (ie, FISH, IHC) to support malignancy in the context of an atypical mesothelial proliferation.
Of the pathologists who responded to our survey, 32 (63%) felt that the clinicians at their institution would treat a patient definitively based on a cytologic diagnosis of MM. This question addressed diagnoses based on both fine-needle aspiration (FNA) specimens as well as effusion specimens without distinction.
We reviewed the medical records of patients with histologically confirmed peritoneal and pleural MM from 2002 through 2011 at Northwestern Memorial Hospital. We identified 20 cases that had a fluid cytology specimen within the 3 months prior to histologic confirmation. Of these 20 cases, 6 (30%) received a cytologic diagnosis of “Positive for malignant mesothelioma” at the time of service based on the cytomorphologic, clinical, and, in some cases, IHC findings. In addition, 6 cases were designated “atypical,” 2 were misclassified as “positive for adenocarcinoma,” 1 was “suspicious for mesothelioma,” and the remainder were classified as benign. There were no false-positive diagnoses of MM in effusion specimens during this time period.
Given the pronouncements of the WHO Blue Book and recent International Mesothelioma Interest Group consensus document, it is interesting to note that the majority of the respondents to our survey were comfortable making a definitive diagnosis of MM in effusion specimens. In arriving at this diagnosis, clinicopathologic correlation and the thoughtful integration of ancillary studies are critical.
The published literature supports the assertion that a cytologic diagnosis of MM supported by clinical data and ancillary studies, although marginal in regard to sensitivity, is extremely specific. Earlier work, much of it prior to the contemporary IHC and molecular paradigm, has documented the sensitivity of effusion cytology for MM to be between 32% and 76%.[7-10] It must be noted that the figures cited by most studies include both “Positive” as well as “Suspicious” results. If only definitive “Positive for mesothelioma” results are counted, as in our case review, the sensitivity is markedly lower. Although not specifically addressed in the prior studies, no instance of a false-positive diagnosis of mesothelioma was noted by any of the authors.
In a frankly malignant effusion where the clinical and/or morphologic features are suggestive of MM, nearly all respondents reported using a panel of IHC stains in order to prove or disprove mesothelial derivation. The utility of IHC in this context is well documented. Although no single stain is entirely sensitive or specific for MM, a panel of stains including both mesothelial (calretinin, WT1, d2-40) and epithelial (Moc31, BerEp4, CEA) markers can often be definitive in the differential diagnosis of adenocarcinoma and MM. Even a minimal 3-antibody panel has been shown to be 96% sensitive and specific in making this distinction.
A smaller proportion of respondents (56%) reported using ancillary studies including IHC and FISH to establish malignancy in proliferations that are morphologically obviously mesothelial, but indeterminate in regard to malignancy. Whereas a number of traditional IHC markers have been evaluated for this purpose (EMA, p53, and desmin), none of them have demonstrated the degree of specificity required to establish an unequivocal diagnosis of MM. More recent work has focused on newer markers such as GLUT1, CD147, and IMP3. Although initial reports of the performance characteristics of these antibodies were promising and demonstrated excellent sensitivity and specificity, more recent publications have failed to confirm these results. It would appear that more data is required before a definitive statement of the utility of these newer IHC markers can be made.
FISH, in contrast, is rapidly emerging as an extremely useful confirmatory tool in the diagnosis of MM. A recent study performed on effusion specimens focused on the evaluation of p16 deletions on chromosome 9 found that 59% of MMs, including 70% of pleural mesotheliomas, demonstrated homozygous p16 deletions. In line with prior work, this study demonstrated the absolute (100%) specificity of a p16 deletion for MM in a mesothelial proliferation. Notably, cytogenetic abnormalities, including homozygous p16 deletions, may be seen in other malignancies as well. As such, it is important to confirm that a proliferation is mesothelial in nature when interpreting FISH results.
The majority (63%) of cytopathologists who responded to our survey felt that the clinicians at their institution would definitively manage a patient on the basis of a positive cytology result. The question specifically referenced both FNA as well as effusion cytology without distinction. An unequivocal diagnosis of MM is useful for therapeutic purposes in many cases, because it expedites therapy and theoretically may avoid seeding of an incision or biopsy site by the tumor. One caveat to the diagnosis of MM via cytology is that there is, as of yet, no reliable way to distinguish the epithelioid, sarcomatoid, and biphasic variants. Although many patients are unresectable at the time of presentation, knowledge of histologic subtype is critical in patients who may clinically be surgical candidates in that a significant sarcomatoid component is considered a contraindication to resection at most centers. Therefore, even with a definitive cytologic diagnosis of mesothelioma and a clinician who accepts it as such, a tissue specimen may still often be necessary to determine subtype.
In conclusion, the results of our survey indicate that a majority of respondents will issue a definitive diagnosis of MM in effusion specimens. This practice appears to be supported by the literature, which documents the excellent specificity of a cytologic diagnosis of MM in the appropriate clinical setting and with supporting ancillary data. Most responding cytopathologists perceive the clinicians at their institution will manage a patient based solely upon a cytologic diagnosis (FNA and/or effusion cytology). Limitations of effusion cytology in this context include a relatively low sensitivity when considering morphology alone as well as an inability to histologically subtype cases.
Table 1. Survey Questions
|What is your practice setting? (n = 55)|
|Academic medical center: 45 (81%)|
|Community hospital: 7 (13%)|
|Military hospital: 2 (4%)|
|Reference laboratory: 1 (2%)|
|Do you make a definitive diagnosis of mesothelioma in cytologic specimens? (n = 55)|
|Yes: 36 (65%)|
|No: 19 (35%)|
|If yes, under what conditions will you make a definitive diagnosis of malignant mesothelioma? (n = 36)|
|Recurrence only: 2 (5%)|
|Primary diagnosis as well as recurrence: 34 (95%) (62% of all respondents)|
|If no, why not? (n = 13)|
|Too much morphologic overlap with reactive mesothelial cells: 6 (46%)|
|Inability to assess stromal invasion in cytologic specimens: 4 (31%)|
|Consensus with clinicians and surgical pathologists: 3 (23%)|
|In which specimen type will you make a definitive diagnosis of malignant mesothelioma? (n = 36)|
|Effusion + Cell Block only: 4 (11%) (9% of respondents)|
|Fine-Needle Aspirate + Cell Block only: 2 (6%)|
|Both: 30 (83%) (55% of all respondents)|
|Do you routinely perform immunohistochemistry (IHC) to support a diagnosis of malignant mesothelioma versus adenocarcinoma? (n = 37)|
|Yes: 35 (95%)|
|No: 2 (5%)|
|Do you perform additional IHC or fluorescence in situ hybridization to confirm malignant mesothelioma versus reactive mesothelial cells? (n = 32)|
|Yes: 18 (56%)|
|No: 14 (44%)|
|In an effusion specimen, which of these features do you find most concerning for malignant mesothelioma? (multiple responses possible, n = 34)|
|High cellularity of mesothelial cells: 16 (47%)|
|Monotonous population of cells without easily identifiable second population of benign mesothelial cells: 13 (38%)|
|Abnormally large clusters of mesothelial cells: 28 (82%)|
|Cytoplasmic density and staining quality: 5 (15%)|
|Cells with mesothelial morphology and marked cytologic atypia: 6 (18%)|
|Do you feel that clinicians at your institution will manage a patient on the basis of a cytologic (fine-needle aspirate or effusion cytology) diagnosis of malignant mesothelioma? (n = 51)|
|Yes: 32 (63%)|
|No: 19 (37%)|
|What is the approximate number of definite mesothelioma or suspicious for mesothelioma cases diagnosed by your laboratory in the past 10 years? (n = 48)|
|<5: 5 (10%)|
|5-15: 10 (21%)|
|16-30: 18 (38%)|
|31-50: 3 (6%)|
|>50: 12 (25%)|