Australian Society of Cytology

Case of the Month

February 2004 - Answer and Discussion

Metastatic adenocarcinoma with a reactive papillomatous mesothelial reaction, origin of the cancer uncertain.

Answer

The pleural fluid, fine needle aspiration and cell block preparations showed a single population of numerous malignant cells in cohesive papillary-like fragments. Papillary aggregates can be a feature of both adenocarcinoma and mesothelioma.

Some fragments had hyaline stromal cores, seen more clearly in the diff quick stains and cell block as a dense pink, amorphous material. These collagen cores have been identified in cases of mesothelioma.

The neoplastic cells had a uniform appearance with high N/C ratio,some increased nucleolar prominence (but no obvious, prominent cherry red nucleoli) and moderate quantities of fine, mildly foamy cytoplasm (no gross vacuolation).

The aspirates showed neoplastic cells in papillary-like arrangements dispersed throughout a background of reactive lymphoid cells.
H+E on the cell block showed similar papillary-like fragments with obvious hyaline material in their cores.
The PASD mucin stain was negative in this case.This does not exclude a diagnosis of adenocarcinoma as a small percentage of tumours do not produce mucin and glycogen degeneration can occur in cells which have been floating in a fluid.

The evidence provided by the routine stains was equivocal. The specimen was classed as a large cell neoplasm with the differential diagnosis being metastatic adenocarcinoma versus mesothelioma.

IMMUNOHISTOCHEMISTRY
Initially the papillary clusters were thought to be a single population of neoplastic cells but, the immune markers clearly showed there to be two distinct cell populations.

One cell population showed strong positive staining for the mesothelial marker Calretinin.

Positive nuclear and cytoplasmic Calretinin staining shows that the cell type is of mesothelial origin, but does not distinguish between reactive and malignant mesothelial cell types.

This result, coupled with a lack of membrane staining for EMA suggested that this population was a reactive mesothelial proliferation as opposed to a malignant process.

The second population was diagnosed as metastatic adenocarcinoma. It showed strong positive staining for EMA .

Note, the EMA stain showed strong cytoplasmic staining and peripheral accentuation due to some membrane staining as opposed to the distinctive, strong membranous positivity in malignant mesothelioma where the staining reaction is located on the microvillous border.

However, the origin of the carcinoma remained uncertain.
CK7 +ve/ CK20 -ve results indicated that the primary tumour might be a lung adenocarcinoma.
PSA stain was negative which excluded a prostatic origin. TTF-1(thyroglobulin transcription factor one) which is expressed in thyroid carcinomas and lung adenocarcinomas was negative, this excluded thyroid as the primary site and made lung unlikely also.

Adenocarcinomas from different sites may have specific morphologies and immunoreactivities, but overlap of features can occur, making distinction of the primary site very difficult as in this case.

Thus, this case left the laboratory as “metastatic adenocarcinoma with a reactive papillomatous mesothelial reaction, origin of the cancer uncertain”.
Further investigation was recommended.

FOLLOW UP
The clinician referred the patient for pleural biopsy and talc pleurodesis.
Operative notes stated that the pleural space was abnormal with multiple nodular lesions present.

The biopsy material consisted of pleural tissue and sub-pleural fat which was infiltrated by a malignant cell population characterised by pleomorphic ovoid nuclei, prominent nucleoli, dense eosinophilic cytoplasm, simple acinar formations and some mucin droplets on mucin stains (PASD).

An extensive panel of immune markers was carried out on this case including a number excluded from the cytology work-up. Strong cytoplasmic immunoreactivity for CAM 5.2, a low molecular weight cytokeratin expressed in the cytoplasm of adenocarcinoma cells, and weak positivity for CD15 which might favour mesothelioma still pointed to a differential diagnosis.

Electron microscopy work was also performed on this case. The malignant cells exhibited microvilli which were short and did not show the characteristic long, thin, intertwined pattern seen in malignant mesothelioma.

Thus, after a detailed histological work-up the results were still not clear cut. Initially, the immunoprofile was equivocal favouring both adenocarcinoma and mesothelioma in various tests.
On the basis of clinical history, morphology, further immune markers and electron microscopy, the final report favoured adenocarcinoma over malignant mesothelioma as in the cytological work.
The primary site still needed to be determined, and at the time of writing, studies were continuing.

Discussion

Metastatic adenocarcinoma is often the major cause of malignant pleural effusions.
Distinguishing between adenocarcinoma and reactive mesothelial cells can be difficult cytologically, particularly in cases of long standing effusion where there may be some reactive mesothelial atypia.
A metastatic adenocarcinoma in a malignant effusion will often have two distinct cell populations - the adenocarcinoma cells and reactive mesothelial cells. Their proportions and architectural presentation will vary dependant on the case.

This case is unusual due to the marked overlap in the morphology of the cell populations.
Routine stains could not differentiate these cell populations initially. The papillary fragments were thought to be one population of malignant cells. Features suggested a differential diagnosis of adenocarcinoma versus mesothelioma.

It was only through the use of immune markers on the cell block material that the lineage of the individual cell types became clear and the true nature of this malignancy identified. It should be noted that the dual cell populations identified cytologically were not observed in the surgical biopsy material. Preservation of architecture and morphological detail has been shown to be excellent in cell blocks, and this may have had some bearing on this result.

Preparation of cell block material from the pleural fluid was instrumental in the diagnosis of this case, by allowing an array of immunocytochemical markers to be used on this specimen. This case highlights the necessity of the preparation of cell blocks from effusion cytology specimens in cases where diagnosis is difficult.

References:

Gray W,McKee G,Diagnostic Cytopathology, 2nd Edition,Churchill Livingstone 2003.
Geisinger K,Raab S,Stanley M,Silverman J,Abati A, Modern Cytopathology, Churchill Livingstone 2004.