Abbreviations: FCC, follicular lymphoma; MCL, mantle cell lymphoma; MZL, marginal zone lymphoma; CLL, chronic lymphocytic leukemia.

Abbreviations: FCC, follicular lymphoma; MCL, mantle cell lymphoma; MZL, marginal zone lymphoma; CLL, chronic lymphocytic leukemia.

The use of panels of markers (CD5, CD23, CD43, cyclin D1, CD10, bcl-6) are used to subclassify the small B-cell lymphomas (46). Some, such as small lymphocytic lymphoma, mantle cell lymphoma, and follicular, have a characteristic immunophenotype (Table 2). For instance, among these markers, the presence of cyclin D1 is considered specific for mantle cell lymphoma (47). Markers such as Ki-67, bcl-6, and CD10 are being investigated as possible prognostic markers in large-B-cell lymphomas (48,49).

Among the most valuable and sometimes the most frustrating IHC markers in the study of B-cell processes are the immunoglobulin light-chain studies. The ability to demonstrate immunoglobulin light-chain restriction is one of the most specific methods to identify neoplastic B-cells. Light-chain restriction can be reliably identified by flow cytometry, if fresh tissue is available. Light-chain restriction can also be easily demonstrated by IHC in plasma cell neoplasms and lymphomas with plasmacytic differentiation. However, inconsistencies and technical problems have beset the identification of light-chain restriction for most other B-cell neoplasms. More recently, studies have shown more consistent results with a modified technique using HIER (50,51).

A series of T-cell markers (CD2, CD3, CD4, CD5, and CD8) is also available for use on paraffin-embedded tissue. Because there is not a good analog to the immunoglobulin light-chain studies in B-cell processes, one must rely on loss of expression of various T-cell markers, altered intensity of expression in the neoplastic cells compared to normal, or alteration of the CD4 and CD8 staining (either marked predominance of one or the other, or T-cells that express both or neither marker). These changes can be suggestive of a T-cell lymphoma, but the final diagnosis of T-cell lymphoma in many cases is dependent on molecular studies to confirm a clonal rearrangement of the T-cell receptor gene (52). As with the B-cell neoplasms, a few T/NK (natural killer)-cell neoplasms, such as anaplastic large-cell lymphoma, have distinct IHC and molecular features.

Hodgkin lymphoma (HL) comprises two distinct groups of diseases: classical HL and nodular lymphocyte predominant HL (NLPHL). Both are characterized by relatively small numbers of neoplastic cells in a background of mixed inflammation in the case of classical HL or lymphocytes in the case of NLPHL. Cells morphologically identical to the large binuclear and mononuclear Reed-Sternberg (RS) cells that typify classical HL can be seen in NLPHL, as well as many forms of non-Hodgkin lymphomas (NHL). Likewise, cells virtually identical to the L&H cells of NLPHL can be identified in other conditions. Immunohistochemistry is essential in distinguishing the two types of Hodgkin lymphoma: the RS cells in classical HL are

CD30 and CD15 positive, whereas the L&H cells in NLPHD are negative for CD15 and CD30 and express CD20. In fact, L&H cells not only show a B-cell phenotype but also show clonal immunoglobulin gene rearrangements (53). With improvements in IHC techniques, even the "classic." immunophenotype of RS is being reevaluated because many cases of HL show at least some CD20-positivity in the neoplastic cells. Additional markers such as BOB.1 and Oct2 (54) or Pax-5 (44), although not 100% specific, can be added to a panel of markers that help distinguish NLPHL from classical HL, or classical HL from its imitators. The immunophenotype of RS cells in a given case greatly depends on technical factors such as fixation and AR (55). Although clonality can be demonstrated in the RS cells with microdissection techniques, the relatively low number of neoplastic cells in cases of HL, means that any clonal bands by PCR would likely to be obscured by reactive cells in the background.

4.3.4. Sarcoma Diagnosis Soft tissue tumors can show a wide variety of histologic appearances, and morphologic overlap is often the rule, rather than the exception. Although there are few specific markers available, panels of antibodies can frequently be used to identify the cell of origin. Some relatively specific markers, including myogenin (skeletal muscle tumors), CD31 (vascular tumors), and CD117 (gastrointestinal stromal tumors), exist, but others, like S-100, react not only with neural-derived tumors but also other tumor types such as melanoma. The coexpression of cytokeratin by a limited number of sarcomas such as synovial sarcoma, chordoma, and epithelioid sarcoma can be a useful feature in subtyping soft tissue tumors, but this can lead to confusion with carcinoma in some cases.

Significant progress has been made in the understanding of the molecular pathogenesis as well as immunohistochemical characteristics of the "small blue cell tumors" of childhood. These include sarcomas such as rhabdomyosarcoma, Ewing sarcoma/primitive neuroectodermal tumor (ES/PNET), desmo-plastic round cell tumor (DRCT), as well as neuroblastoma and acute lymphoblastic lymphomas (ALL). Morphologic overlap is very common among these tumors, whereas prognosis and therapy are becoming more individualized. Molecular pathology has elucidated characteristic translocations in some of these sarcomas [alveolar rhabdomyosarcoma (56), ES/PNET (57), DRCT (58)] and molecular techniques such as reverse transcription (RT)-PCR are able to identify these abnormalities even in formalin-fixed, paraffin-embedded tissue (59). However, these techniques are available only in specialized centers, and the majority of pathologists rely on IHC panels for the diagnosis (Table 3).

In the case of markers of cell lineage, desmin, myoD1, and myogenin indicate muscle differentiation (60). Desmin is a highly sensitive marker for rhabdomyosarcoma, but it lacks specificity. MyoD1 and myogenin are muscle-restricted nuclear transcription factors, highly specific for skeletal muscle differentiation and, therefore, for rhabdomyosarcomas (61,62). Among them, myogenin is the preferred marker (Fig. 3) because myoD1 usually shows more nonspecific background staining.

The ES/PNETs are positive for CD99 and for the more recently introduced marker FLI-1. Although CD99, the product

Table 3

IHC Panel in Small Blue Cell Tumors

Desmin Myogenin CD99 CD43 NF FLI-1 WT-1

Abbreviations: MS, rhabdomyosarcoma; EW, Ewing sarcoma; DRCT, desmoplastic small round cell tumor; NB, neuroblastoma; LL, lymphoblastic lymphoma.

Fig. 3. Diffuse nuclear staining for myogenin in alveolar rhabdomyosarcoma.

of the MIC2 gene, is very sensitive for ES/PNET (63), it is not as specific as it was initially thought because it has been documented in other small-cell sarcomas such as rhabdomyosarcoma, synovial sarcoma, and ALL. Recently, overexpression of the nuclear protein FLI-1 has been demonstrated in about 70% of ES/PNET (64). The majority of ES/PNET cases show the translocation t(11;22)(q24;q12) that leads to production of the EWS-FLI-1 fusion protein, which is a transcription activator with transforming potential. Whereas the EWS/FLI-1 chimeric gene is specific for ES/PNET, the IHC antibody that recognizes FLI-1 is not. Positive staining for FLI-1 has not been seen in other small-cell sarcomas, but has been documented in ALL and some vascular tumors.

Another small round blue cell tumor with a distinctive phe-notype is DRCT. It shows coexpression of cytokeratin, vimentin, desmin, and neuron-specific enolase (65). This tumor has a characteristic translocation, t(11;22)(q13;q12), which gives rise to the production of the chimeric EWS/WT-1 protein. Antibodies against WT-1 are used to detect this protein by IHC (66). This protein is not entirely unique to DRCT as it is also expressed in Wilms tumor. However, among the other small round blue cell tumors, its presence is specific for DRCT.

4.3.5. Tumor Prognostic Factors Although IHC is predominantly used to determine tumor lineage or highlight tumor cells, it is becoming increasingly evident that some IHC markers have prognostic value and even direct therapeutic implications (67). Among these are ER/PR and HER-2, which are routinely evaluated by IHC to help direct therapy in breast carcinoma. The expression of CD117 (c-Kit) helps separate gastrointestinal stromal tumors (GISTs), which respond well to imatinib, a tyrosine kinase inhibitor, from other mesenchymal tumors of the gastrointestinal (GI) tract.

p53 expression is another commonly used prognostic marker (68,69). The p53 gene, located on chromosome 17p, encodes a nuclear DNA-binding protein that regulates cell growth. Mutant forms of the p53 protein are resistant to degradation and accumulate in the nucleus, where it is detectable by IHC. In the normal state, p53 is rapidly degraded and is usually not detectable. Overexpression or accumulation of this protein is associated with a poor prognosis in a wide variety of tumors. Studies have also shown that tumors harboring mutations in p53 might be more resistant to chemotherapy and radiation therapy.

4.3.6. Tissue Microarrays A significant technical advancement is the recent introduction of the tissue microarray (TMA) method. This technique involves harvesting small 0.6-mm cores of tissue from hundreds of archival paraffin blocks and constructing a single TMA block, which can be used not only for IHC but also for in situ hybridization techniques. With this technique, one can examine many examples of a given tumor or group of tumors at one time using a minimum of resources. It is a highly efficient method to that allows for fast, parallel immunohistochemical profiling of large numbers of cases and is an excellent way to evaluate new protein markers. It can serve as a powerful quality assurance tool in assessing the specificity and sensitivity of different antibodies (70). In addition, its use in the research laboratory might contribute to the expeditious translation of new molecular findings into clinical applications.

4.4. MOLECULAR/IMMUNOHISTOCHEMICAL CORRELATION The following are common examples in which genetic events translate into phenotypic changes that can be detected by immunohistochemistry (71-73). In some cases, these changes are best detected by molecular means such as fluorescence in situ hybridization (FISH) or PCR, whereas in others, IHC is the method of choice. In many instances, the use of IHC is simply an easier and more accessible laboratory test that gives virtually the same information. In other areas, IHC serves well as a screening test, whereas the PCR or in situ hybridization studies play more of a confirmatory role.

4.4.1. Genetic Alteration Resulting in Overexpression of a Protein Her-2/Neu in Breast Carcinoma The field of IHC has been fruitful in the search for independent prognostic indicators in breast carcinoma. The detection of ER and/or PR in the tumor cells predicts response to hormone-related therapy and a better outcome. In contrast, overexpression of HER-2, an oncogene in the epidermal growth factor receptor family, has been associated with poorer outcome with respect to disease-free status and overall survival (74). On the other hand, some studies have shown that such patients benefit from doxorubicin therapy and the humanized monoclonal antibody to the HER-2 protein, trastuzumab. Although there is agreement that this is an important parameter to be assessed in patients with breast carcinoma, there is considerable debate as to the most appropriate way to evaluate tumors for this overexpression.

Although immunohistochemistry is a quick, inexpensive method for evaluation of HER-2 overexpression, a lack of standardization of techniques, use of multiple antibodies, and variable scoring have plagued the assessment of HER-2 by IHC (75). With standardized protocols for the assay and the evaluation of results, reproducibility is much improved. A number of preanalytic factors can alter results, including difference in type and time of fixation and selection of appropriate control and test material. The scoring of HER-2 IHC ranges from 0 (no membrane staining in invasive tumor) to 3+ (strong, complete membrane staining) (76). Practice with the standard scoring system should reduce variation. Most studies demonstrate a close correlation between strong positivity (3+) by IHC (Fig. 1B) and amplification of the HER-2 gene by FISH studies. There is less agreement when the IHC results are intermediate, although the large majority of 1+ and 2+ results by IHC are shown to be negative for amplification by FISH. FISH and, more recently, chromogenic in situ hybridization (CISH) appear to work well in archived formalin-fixed tissue. These assays are typically more costly and time-consuming than IHC, but the level of variability seen in IHC studies has prompted some to advocate FISH as the method of choice for screening patients for Her2/neu overexpression (77). An advantage of FISH is that it allows one to determine in part the mechanism by which Her2lneu is overexpressed. It has been shown that the majority of cases of overexpression are caused by gene amplification, and this technique permits the observer to quantify that amplification by actually counting the number of copies of the Her2lneu gene present in the tumor cells. Processing with enzymatic digestion can lead to either no signal or to autofluorescence that might obscure signal, leading to reports in the literature of false-negative rates up to 10%. In addition, slide preparation leads to distortion of tissue morphology such that it might be difficult to distinguish between invasive and in situ components of tumor.

A recent symposium offered by the College of American Pathologists offered a "consensus testing algorithm" (78) for evaluation of Her2lneu status in breast carcinomas. They suggest that if a lab's concordance between FISH and IHC studies is high enough (>90% for scores of 3+ and 0), then IHC can be used as a screening test with reflex testing by FISH for cases that show 1+ or 2+ reactivity by IHC. If the concordance rates are lower than 90%, the laboratory should consider running all tests by FISH. c-Kit Overexpression in Gastrointestinal Stromal Tumors The gastrointestinal stromal tumor (GIST) is the most common mesenchymal neoplasm of the gastrointestinal (GI) tract. This entity has been a source of confusion in terms of classification for many years because it has an ability to show differentiation along several different lines such as myogenic (smooth muscle), neural, and bidirectional differentiation. It was shown that GISTs originate from the interstitial cells of Cajal, also known as GI pacemaker cells, which have immunophenotypic and ultrastructural characteristics of both smooth muscle and neuronal differentiation in varying degrees and serve to regulate peristalsis. They are characterized by expression of c-Kit (CD117) in virtually all cases (Fig. 4), CD34 in over 70%, smooth muscle actin in less than 30%, and, less commonly, S-100 protein and desmin (79).

c-Kit is a transmembrane tyrosine kinase receptor that is expressed at high levels in hematopoietic stem cells, mast cells, melanocytic cells, germ cells, and the interstitial cells of Cajal (ICC). In 1998, Hirota and colleges reported that some GISTs contain an exon 11 mutation in the c-Kit proto-oncogene (80). This mutation in the c-Kit gene leads to overexpression of the tyrosine kinase moiety for the c-kit protein (CD117), which is recognized as a reliable phenotypic marker for this neoplasm. Mutations that result in c-Kit overexpression are thought to play a major role in the pathogenesis of GIST. The recent introduction of a receptor tyrosin kinase inhibitor (STI-571, ima-tiniber or Gleevec) that inhibits activated c-Kit protein, for the first time provided an effective treatment for recurrent or metastatic GIST (81). Expression of CD117 is a reproducible diagnostic criterion for GIST and was a requirement for many clinical trials involving STI-571, although recent data suggest that even in the absence of c-Kit staining by IHC, GISTs might respond to STI-571. It has been proposed that the term GIST be applied only to tumors expressing CD117, although rare exceptions might exist that include lesions that appear immunohistologically nonreactive secondary to poor fixation and preparation, are c-Kit negative because of sampling error, or have ceased to express c-Kit because of some form of clonal

evolution (STI-571 therapy). Antibodies to c-Kit typically show diffuse cytoplasmic and/or membranous staining, with rare tumors showing perinuclear staining (82). Mast cells serve well as a positive internal control to confirm immunoreactivity of the tissue and that the stain is working properly. In less than 2% of cases, an otherwise typical tumor lacks c-Kit overexpression and should probably be labeled as spindle cell (or epithe-lioid) stromal neoplasm most consistent with GIST (80).

It should be noted that other tumors that involve the GI tract might also express c-kit. Melanoma, seminoma, granulocytic sarcoma, as well as other malignant spindle cell tumors have been reported to be positive for CD117. For this reason, the expression of CD117 must always be interpreted in the context of the H&E morphology and clinical findings.

4.4.2. Genetic Mutations Resulting in Loss of Protein Expression E-cadherin in Lobular Breast Carcinoma

E-Cadherin is a cell surface glycoprotein involved in cell adhesion. The protein is encoded by the CAD1 gene on chromosome 16q and loss of its expression is associated with increased invasiveness and higher-tumor grade (83). Dysregulation of E-cadherin is seen in many carcinomas and is caused by a heterogenous mix of genetic mutations that can be detected by molecular methods (84). Mutations of the E-cadherin gene are particularly prevalent in poorly cohesive neoplasms like lobular carcinoma of breast and gastric carcinoma (85). Of interest, mutations of this gene are rarely seen in ductal or medullary carcinoma of the breast. Because of this dichotomy, IHC can be used to separate lobular and ductal breast carcinomas that have ambiguous H&E morphology. Lack of staining with antibodies to E-cadherin in an in situ or invasive breast carcinoma indicates lobular rather than ductal carcinoma.

4.4.3. Mismatch Repair Gene Expression in Colonic Carcinoma There are two recognized genetic pathways for the development of colonic carcinoma. The less common of the two is seen in about 10-15% of sporadic cases of colon cancer and all cases of hereditary nonpolyposis colon cancer syndrome (HNPCC) or Lynch syndrome (86). Over 90% of these cases show mutations in the hMLH1 and hMSH2 genes whose products are involved in DNA mismatch repair, a critical proofreading function in DNA replication. A germline defect in one of these genes carries with it a lifetime risk for colon cancer of about 80%. The result of such alterations is microsatellite instability (MSI), which is the molecular hallmark of this group of tumors. The clinical relevance for identifying this group of patients is that they have a better survival rate but have a higher incidence of metachronous tumors. In addition, it allows for earlier screening of potentially affected relatives. Although molecular studies for alterations in these genes are available, tumors can be screened in a technically easier and less costly way with IHC for hMLH1 or hMSH2 proteins. Loss of expression of these gene products is an effective surrogate for the molecular presence of microsatellite instability (87,88).

4.4.4. Translocations Result in Expression of a Chimeric Protein or Activation of a Normal Gene with Overexpression of a Normal Protein ALK1 in Anaplastic Large Cell Lymphoma One entity that has been defined by IHC is the CD30-positive anaplastic large-cell lymphoma (ALCL) (89). The classic form of ALCL has sheets of large "hallmark" cells with vesicular embryoid or reniform nuclei, one or multiple nucleoli, abundant amphophilic cytoplasm, and a perinuclear hof, Although many morphologic subtypes exist, they have in common the expression of CD30 in a membrane and Golgi distribution (Fig. 5A). A subset of these tumors were shown to have a characteristic t(2;5)(p23;q35) chromosomal translocation that fuses the ALK and NPM genes and leads to production of a novel NPM-ALK fusion protein containing the N-terminal portion of nucleophos-min and the cytoplasmic domain of ALK, a neural-associated receptor tyrosine kinase (90). It has not been possible to predict which tumors will have abnormal ALK expression on the basis of morphology.

Fig. 5. (A) CD30 highlights neoplastic cells in ALCL; (B) ALK-positive neoplastic cells in ALCL, with cytoplasmic and nuclear staining.

The laboratory that initially described the p80 protein (NPM-ALK) first raised polyclonal antibodies to the tyrosine kinase domain (91). Since that time, purified monoclonal antibodies have been produced that are used for routine IHC analysis. This marker is quite specific in that it marks no other type of lymphoma and is absent from normal tissues except for weak posi-tivity in a small subset of cells in the brain. A few nonlymphoid entities have been shown to be positive for ALK1, including inflammatory myofibroblastic tumor and rare soft tissue tumors. In cases of ALCL, the IHC staining pattern gives an idea about the underlying translocation involved. The classic NPM-ALK fusion protein shows pattern of nuclear and cytoplasmic staining because although ALK is present mainly in the cytoplasm, wild-type NPM and the NPM-ALK fusion protein can form dimers localized to the nucleus (Fig. 5B). When translocation partners other than NPM are present, the staining is typically limited to cytoplasm or to the cell membrane. ALK expression by the neoplastic cells is a favorable prognostic factor, but the specific genetic rearrangement leading to that overexpression does not seem to correlate with outcome (92).

The translocations in ALCL occur over large areas within the ALK and NPM, thereby limiting the use of standard PCR as a diagnostic tool. Furthermore, up to 20% of cases have translocation partners other than NPM, which would not be detected even with RT-PCR. FISH studies can detect either the specific t(2;5) or can evaluate for ALK rearrangements, allowing detection of variant translocations involving the ALK gene. Given the fact that IHC is less expensive, faster, and easier than the molecular methods and the fact that the pattern of staining can predict variant translocations, IHC testing is favored for evaluation of ALCL (93). Cyclin D1 in Mantle Cell Lymphoma Overexpression of cyclin D1, a nuclear protein whose expression permits the cell to transition from the G1-phase to the

S-phase in the cell cycle, is a defining feature of mantle cell lymphoma. In this lymphoma, a characteristic chromosomal rearrangement, t(11;14)(q13;q32), juxtaposes the bcl-1 (CCND1 gene) locus at 11q13 with the IgH locus at 14q32. Although increased cyclin D1 expression is the hallmark of mantle cell lymphoma, it is also reported in multiple myeloma and hairy cell leukemia (72). Detection of cyclin D1 by IHC has been hindered by technical difficulties (94). In some instances, cyclin D1 staining can show nonspecific cytoplasmic staining (caused by biotin) that is clearly distinct from the mosaic pattern of nuclear staining that characterizes a positive reaction (Fig. 1A). In negative cases, one can assess the internal control seen as positive nuclear staining in scattered endothelial cells. With careful preparation using a high-temperature, high-pH AR method, it can be detected in over 95% of cases of mantle cell lymphoma (95). By contrast, PCR methods on paraffin-embedded tissue detect the t(11;14) rearrangement in 50-70% of cases of mantle cell lymphoma. More recent reports using FISH give a detection rate close to 100% (96,97). This FISH procedure can be applied to formalin-fixed, paraffin-embedded tissue as well as unfixed cells.

Similar mechanisms of overexpression of regulatory proteins produce other lymphomas. For instance, in follicular lymphoma, the antiapoptosis protein bcl-2 is commonly overex-pressed because of translocation of the bcl-2 gene on chromosome 18 is translocated to the IgH locus on chromosome 14, leading to overexpression. This overexpression blocks normal apoptosis, leading to overgrowth of follicle center cells.

Was this article helpful?

0 0
How To Prevent Skin Cancer

How To Prevent Skin Cancer

Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.

Get My Free Ebook

Post a comment