College of American Pathologists
Printable Version

  Cover Story


cap today

Special stains and immunostains

Histology labs eye new kids on the block

October 2003
William Check, PhD

To demonstrate the value of immunostains, Allen M. Gown, MD, medical director and chief pathologist at PhenoPath Laboratories, Seattle, goes back to his days as a resident in the 1970s.

“When we worked up a case and H&E didn’t give the final answer, we first elected special stains,” he says. That included things like reticulum stain. Now when a diagnostic dilemma arises, in most laboratories the first step is to think of immunostains. “Antibodies to type IV collagen can be used routinely to address many of the same issues that were addressed by reticulum stains,” Dr. Gown says. Looking for organisms might be one place where traditional stains prevail. “Even then sometimes immunostains do a good or even better job,” he says, citing Steiner stain for Helicobacter pylori.

Dr. Gown prefers to distinguish immunostains from special stains. “Maybe because of the way immunostains were introduced into the laboratory, in some people’s minds they are a kind of special stain,” he says. In his view, special stains are empirical—they just happen to work—whereas immunostains are defined by an antibody of known specificity.

Because they are based on antibodies, immunostains have proved more of a challenge to automate than special stains. Immunostains require special considerations in tissue preservation and preparation to optimize their use. “They are less forgiving than special stains,” Dr. Gown says. He cites antigen retrieval as something generally required for immunostains but not for special stains.

Forefront immunostains include antibodies to relatively esoteric tumor-specific markers, such as the FLI-1 gene product found in Ewing’s sarcoma and the WT-1 gene product that characterizes desmoplastic small cell tumor. Both of these gene products represent C’ terminal elements of chimeric fusion proteins created by tumor-specific chromosomal translocations.

For such tumor-type specific markers, says Paul Swanson, MD, director of anatomic pathology at the University of Washington Medical Center, Seattle, “when the translocation-defined portion of the protein is expressed in the appropriate histologic context, it can be a reliable marker for that disease.” In these cases, immunohistochemistry detects what otherwise would be identified by cytogenetic analysis or rtPCR for the chimeric fusion transcript. “We cannot claim either sensitivity or specificity benefits with immunostains relative to cytogenetics,” Dr. Swanson says. “But sometimes because of tissue handling, conventional cytogenetics or molecular techniques cannot be performed. In particular, rtPCR is extremely dependent on fresh or properly prepared tissue,” Dr. Swanson says. If tissue is formalin-fixed and routinely processed, it can be hard to retrieve intact mRNA.

In these situations, IHC offers an alternative. Immunostains that recognize lineage-restricted markers composed of a family of transcription regulators have also appeared on the forefront of diagnostic pathology. Most widely used among these is thyroid transcription factor 1, or TTF 1, says Dr. Swanson. “It is associated with elements of foregut development in fetal tissues,” he says. It is also expressed in differentiated thyroidal and pulmonary epithelium, and as such is a very sensitive and reasonably specific marker for thyroid and pulmonary glandular and endocrine neoplasms. It also has limited value in extrapulmonary neuroendocrine carcinoma.

A related marker, CDX-2, is preferentially expressed in midgut and hindgut epithelium. “It is a reasonably specific and sensitive marker for gastrointestinal carcinomas, especially carcinoma of the colon and rectum,” Dr. Swanson says.

TTF-1 and CDX-2 are particularly relevant to the diagnosis of metastatic carcinoma, especially in those cases where primary site of involvement is uncertain or unknown (occult clinical disease). “A lot of IHC today revolves around that broader question,” Dr. Swanson says.

Other recent markers purport to distinguish between noninvasive proliferative lesions and overtly invasive malignant disease. High-molecular-weight keratins have been joined by the p53 analogue p63 as markers of basal cells in prostate. P63 also complements antibodies to smooth muscle determinants as a marker of myoepithelial cells in breast and other organs. Immunoreactivity for these markers in the appropriate context is evidence of an intact or attenuated basal cell or myoepithelial layer (non-neoplastic or in situ disease), whereas the absence of these markers is presumptive evidence of extension of carcinoma beyond normal structures.

In contrast, α-methyl acyl coA racemase (p504s) is selectively expressed in malignant epithelium of prostate cancer, ostensibly providing a positive marker for carcinoma instead of the traditional inference of malignancy from the loss of normal elements. Dr Swanson calls it “somewhat controversial.”

Still other markers of increasing relevance to diagnostic pathology are those, such as Her2/neu, topoisomerase–2α, and thymidylate synthetase, that predict response to specific classes of chemotherapeutic agents, a role until recently played only by estrogen and progesterone receptor proteins.

“For me, the important question posed by these markers is not whether a selective panel of chemotherapeutic agents or other treatment options is available to the patient,” Dr. Swanson says, “but whether such treatment is of proven or expected clinical value.” In a similar vein, one might question if there is a defined benefit to the physician and the patient in making an immunohistochemical determination of the origin of metastatic disease.

“In most cases there is,” Dr. Swanson says, “but not always. What we detect by immunohistochemistry may help dictate management, but will it reliably predict outcome?”

William Check is a medical writer in Wilmette, Ill.