One week was just long enough for presentation of the material that was covered in the scientific sessions at UltraPath VIII (Table 1). However, assimilation of the wealth of information in these sessions will require a much longer time. The theme inherent theme of the scientific sessions was that of improved accuracy in the diagnosis of neoplastic and non-neoplastic lesions. Improvement in diagnostic accuracy is achieved through the integration of the results of molecular techniques as well as other ancillary studies with the morphologic information provided by electron microscopy. Here are some of the theme related pearls that were either presented or discussed in the scientific sessions.
|
Although a variety of immunohistochemical stains are often used in the diagnosis of these tumors, staining patterns can be quite variable. Tissue fixation, antibody concentration and number of neurosecretory granules are some factors that affect the outcome of immunohistochemical staining. Staining for chromogranin is absent in gastric and rectal carcinoids as well as most small cell carcinomas. Paragangliomas stain poorly for chromogranin. S-100 protein stains the sustentacular cells in paragangliomas. This immunohistochemical reaction can prove useful in the diagnosis of paragangliomas that occur in unusual locations such as the larynx and mediastinum. Synapatophysin is poorly expressed in paraffin embedded neuroendocrine and neuroepithelial tumors. Merkel Cell tumors exhibit a punctate, dotlike staining for ctyokeratin. However, in B5-fixed Merkel cell tumors, the cytoplasmic staining for cytokeratin is diffuse. Staining for low molecular weight cytokeratin can be used to differentiate neural (neuroendocrine) and epithelial (neuroepithelial) type neuroendocrine tumors since staining for low molecular weight cytokeratin is absent in the neural type tumors. Protein Gene Product (PGP 9.5), a brain protein that is distinct from neuron specific enolase, is diffusely distributed in the cytoplasm of many neuroendocrine tumors except those from the GI Tract.
Given the variation in the outcome of immunohistochemical staining, electron microscopy retains its key role in the diagnosis of neuroendocrine and neuroepithelial tumors, especially the poorly differentiated neuroendocrine neoplasm. The diagnostic organelle is the dense core secretory granule that has a diameter range of 80 to 450 nanometers. Silver and uranaffin staining of thin sections confirm and distinguish these neuroendocrine granules from other electron dense, intracytoplasmic structures such as lysosomes and zymogen granules. Electron microscopy also provides information about the range of ultrastructural features that can be encountered in neuroendocrine tumors. Cytoplasmic processes that contain neurofilaments are usually present in neural type, neural crest derived neuroendocrine tumors. The neuroepithelial type tumors will exhibit additional features of epithelial differentiation such as cell junctions, tonofilament bundles and gland formation.
However, ultrastructural and immunohistochemical identification of neuroendocrine differentiation does not always equate with a diagnosis of neuroendocrine neoplasm. Focal neuroendocrine differentiation may be present in large cell carcinomas and prostate adenocarcinomas. Thus an integrated approach to the ultrastructural and immunohistochemical staining pattern of these tumors is strongly emphasized.
Metastatic tumors in the liver are a diagnostic challenge. Ultrastructural examination of these lesions will demonstrate a mixture of cellular organelles. Large, intracytoplasmic dense bodies, a feature of primary hepatocellular carcinomas may also be found in tumors that metastasize to the liver. Therefore, there is no single organelle that will differentiate primary liver cells carcinomas from secondary, metastatic tumors to the liver. Another diagnostic pitfall is the presence of rare intracytoplasmic lumina in metastatic adrenal carcinomas. When interpreted in isolation, this ultrastructural feature can result in the erroneous diagnosis of either a metastatic poorly differentiated adenocarcinoma or a poorly differentiated primary liver carcinoma.
The immunohistochemical staining reaction for liver cell carcinomas is sometimes helpful. Cross reactivity between biliary glycoprotein and polyclonal CEA is a helpful in establishing a diagnosis of hepatocellular carcinoma. Positive staining of a neoplasm in the liver for HMB-45, is not an absolute criterion for making the diagnosis of metastatic melanoma. Hepatoblastomas can be HMB-45 positive and it seems that this positive staining is probably present even with the HMB-45 antibodies that are specific for melanosomes. Thus accurate diagnosis of primary and metastatic liver tumors requires the consideration of the results of all available studies as well as correlation of these studies with the clinical information.
Soft Tissue Tumors
Immunohistochemistry and electron microscopy are still jockeying for the role of the better ancillary study for the typing of soft tissue tumors. Although the utility of immunohistochemical staining patterns cannot be denied, there are several types of soft tissue sarcomas in which ultrastructural morphology is better than the information that is obtained from immunohistochemical staining (Table 2).
|
Distinction of sarcomas from other tumors sometimes hinges on the absence of cytokeratin staining in soft tissue sarcomas. However, staining for cytokeratins 8, 18 and possibly 19 can be exhibited by soft tissue sarcomas. PCR studies suggest that cytokeratin production is an inherent feature of cells of these types of sarcomas.
Variation in immunohistochemical staining is also a property of soft tissue tumors. False negative staining for epithelial membrane antigen (EMA) can occur in perineuriomas whose diagnosis is based on ultrastructural examination. The false negative staining for EMA is a result of low antibody concentration.
Kidney
Ultrastructural examination of renal transplant biopsies is emphasized since it can be used to detect other pathologies that may coexist with the light microscopic features of rejection. The criteria for selection of a transplant biopsy for electron microscopy are: clinical evidence of proteinuria, an unusual rejection pattern such as unexplained renal failure and discordance between the morphologic features of consecutive renal biopsies. EM may demonstrate features that suggest coexistent lesions such as either transplant glomerulopathy or membranous glomerulonephritis. Transplant glomerulopathy is a de-novo glomerulonephritis that affects the renal transplant. Ultrastructural examination of tissue form the transplant kidney demonstrates an absence of immune complexes widening of the subendothelial region in the glomerular tuft. Cost Analysis
Although the cost of performing a single immunohistochemical stain on a tissue sample is cheaper than doing ultrastructural examination of this tissue, satisfactory immunohistochemical evaluation of a variety of neoplastic and non-neoplastic lesions requires the use of a panel of immunohistochemical antibodies. Depending on the number of antibodies in this panel, the cost of the total immunohistochemical study can exceed the cost for ultrastructural examination. If electron microscopy is used as the initial ancillary study, it can yield a diagnosis that eliminates the need for immunohistochemical staining. When electron microscopy fails to yield a diagnosis, the ultrastructural features of the lesion will suggest a more restricted panel of immunohistochemical antibodies that will provide a diagnosis.
Ancillary Studies
Choice of an ancillary study for the evaluation of neoplastic and non-neoplastic lesions should be based on the quality of the information that the study will provide as well as whether this information will lead to an accurate diagnosis. If the ancillary study will be performed at another medical center, select a reputable site for the performance of the test. Be prepared to select additional studies if an initial technique does not work.
Cytogenetics is an excellent technique for the investigation of soft tissue neoplasms as well as small round cell tumors of children. However, this is a slow technique that does not always work. The pitfalls of cytogenetic studies are that: discrepant results are generated in different laboratories; 80% of all human cancers lack a specific chromosomal alteration and many of the diagnostic cytogenetic alterations are based on evaluation of single cases rather than studies of groups of similar neoplasms.
Comment
This selection of pearls from the scientific session is quite
incomplete because of the limited space available in the newsletter.
If you did not attend UltraPath VIII, you will not be shortchanged
because the proceedings from the scientific sessions will be published
in futures issues of Ultrastructural Pathology.
The Social Program