key: cord-0044592-ohvo9mb8
authors: Lester, Susan C.
title: Special Studies
date: 2010-12-27
journal: Manual of Surgical Pathology
DOI: 10.1016/b978-0-323-06516-0.10007-4
sha: 58d89708e070db6199d8675a4e232c8b574e62d3
doc_id: 44592
cord_uid: ohvo9mb8

nan

The pathologist's H&E is like the clinician's H&P (history and physical) -basic exams to be performed on every patient or specimen forming the cornerstone of diagnosis. However, the pathologist is no longer limited to the H&E; there are a wide variety of special studies available to evaluate pathologic processes, from simple histochemical stains to global gene expression patterns. Pathologists are now clinical cell biologists. Familiarity with the types of special studies available is important as the initial processing of the gross specimen may limit the types of studies that can be performed.

Almost all histochemical stains are suitable for formalinfixed tissues. Common stains and their uses are listed in Table 7 -1. However, numerous other types of stains and modifications are used and pathologists must be aware of individual laboratory practices.

The WebPath section of the University of Utah site (http://library.med.utah.edu/webpath) has useful descriptions of special stains and illustrative photographs. 

The development of methods to detect antigens on tissue sections with antibodies was a major advance in surgical pathology. Immunohistochemical (IHC) studies are most frequently used for the following purposes:

• Classification of tumors (e.g., carcinoma vs. lymphoma, B cell vs. T cell lymphoma). • Identification of in situ lesions vs. invasive carcinomas (e.g., myoepithelial markers in breast cancers, basal cell markers in prostate). • Prognostic factors (e.g., Ki-67 in glioblastomas).

• Predictive factors to guide specific therapy (e.g., c-KIT, estrogen and progesterone receptors, HER2/neu). • Identification of extracellular material (e.g., β-2 microglobulin amyloid). • Identification of infectious agents (e.g., CMV or HSV).

Use of Immunohistochemistry. A differential diagnosis is generated after examination of the H&E-stained slides. IHC is then used to gain evidence for or against diagnostic possibilities. "Trolling" cases through an immunohistochemistry laboratory by ordering numerous antibody studies without a clear reason in mind is more likely to lead to misguided diagnosis due to aberrant immunoreactivity than to provide an unexpected correct diagnosis.

Panels. There are no absolute rules for immunoreactivity in cells and tissues. Aberrant immunoreactivity or loss of immunoreactivity is occasionally observed for all antibodies, either due to biologic variability (e.g., occasional keratin-positive melanomas) or technical factors (e.g., impure antibodies, cross-reaction with other antigens, failure to preserve antigenicity). Thus, immunohistochemical markers are used most effectively as panels of markers with interpretation based on an immunohistochemical profile.

Slides for Immunohistochemistry. Tissue is often dislodged from normal glass slides during the treatments required for IHC. Thus, slides must be coated (e.g., with glue, poly-L-lysine, gelatin, albumin) or special commercial slides must be used. If slides are being prepared by another laboratory, the type of glass slide to be used must be specified.

Factors Affecting Immunogenicity. There are numerous variables that can affect antigenicity. The most common are listed below. Each laboratory must optimize its procedures for each antibody used. Studies on tissues or slides not prepared in the routine fashion for a laboratory must be interpreted with caution.

• Type of fixative. Some fixatives destroy some antigens (e.g., Bouin's diminishes ER immunoreactivity, keratins are not well preserved in B5). 1 Most studies are based on formalin-fixed tissue. Results cannot be assumed to be equivalent for other fixatives. • Length of time of fixation in formalin causes protein cross-linking, and antigenicity generally decreases with fixation times over 24 hours. To some extent, this effect can be reversed using antigen retrieval methods. Antigenicity can also be reduced if the tissue is fixed for too short a period (e.g., less than 6 hours). • Prior decalcification in hydrochloric acid. Decreases antigenicity of some epitopes (predominantly nuclear) but not others (predominantly cytoplasmic). 2 Decalcifying agents using EDTA did not alter immunogenicity.

• Decreased: estrogen receptor (ER), progesterone receptor (PR), p53, BerEp4 (tumor cells) , H blood group. • Not affected: calcitonin, chromogranin, GCDFP-15, HMB 45, thyroglobulin, S100, prostate-specific antigen (PSA), keratins (CK 20, CAM5.2, AE1/AE3), A and B blood groups, others. • Temperature of baking the slide. • Length of time since the glass slide was cut. The immunoreactivity of the majority of antigens declines over days to weeks with potential complete loss at one month. 3, 4 The loss may be due to exposure of tissue to air with oxidation of amino acids, as the immunogenicity of tissue deeper in the block can be preserved for many years. Antigen retrieval methods do not completely restore the antigenicity of old slides. Coating slides with paraffin, storing the slides in a nitrogen desiccator, and/or storing at lower temperatures can partially preserve antigenicity. However, studies should be performed on newly cut slides, if possible. • Antigen retrieval procedures (e.g., proteolysis, heating [microwave, steam] , special incubation fluids). To some extent these methods reverse the effects of formalin fixation. Variable effects are observed for different antibodies. • Type of antibody (polyclonal vs. monoclonal vs. mixture of different monoclonals, epitope detected, mouse vs. rabbit). Very different results can be obtained with different antibodies to the same protein or different commercial sources of the same antibody. • Incubation time, incubation temperature, dilution of antibody. • Methods of signal amplification.

Controls are essential for the appropriate interpretation of immunohistochemical studies and ensure that all steps of this complicated procedure have been performed adequately.

Positive controls consisting of tissues known to be immunoreactive should be included each time an antibody is used for a test case. Internal positive controls should always be evaluated when present, as they control not only for the technique used but also for the antigenicity of the tissue under investigation. The immunoperoxidase table lists normal cells that are generally immunoreactive for each antibody. Some laboratories have used vimentin as a control for immunogenicity as almost all tissue should demonstrate positivity. 5 Given the wide and non-specific distribution of vimentin, smooth muscle alpha actin may be more useful in this context as pericytes, vascular smooth muscle, and myoepithelial cells present in most tissues are immunoreactive.

Examples of internal controls:

• S100: Normal nerves, melanocytes, and Langerhans cells in epidermis, cartilage, some myoepithelial cells, skin adnexa • Estrogen and progesterone receptors: Normal luminal cells in ducts and lobules of the breast Negative controls usually consist of replacing the primary antibody with non-immune animal serum diluted to the same concentration as the primary antibody. No positive reaction should be present. If multiple primary antibodies are used reactive with different target antigens, then they may serve as negative controls for each other. Although the best negative control would be to use antibody preabsorbed against the target antigen, this is rarely practical in a diagnostic laboratory. Diagnostic slides should also be evaluated for internal negative controls. Aberrant immunoreactivity of tissues that should not be positive is indicative that the immunoreactivity is nonspecific and the study should not be used for interpretation.

The following features must be taken into consideration when evaluating studies:

Location of Immunoreactivity. Antigens are present in specific sites. Some antigens may be present in more than one location or be extracellular.

Nonspecific positivity should be suspected when immunoreactivity is present in atypical locations:

• Background: Suspect nonspecific positivity if normal cells or noncellular components are positive. This can occur with suboptimal performance of the assay or suboptimal antibodies. • Edge artifact: Antibodies can pool at edges or holes in tissue. True positivity should also be present in the center of the tissue. • Necrosis or crushing of cells: Nonspecific positivity can be seen in disrupted cells. Although keratin is generally reliable in necrotic tumors, other markers generally should not be interpreted. • Inappropriate location (e.g., cytoplasm instead of nucleus):

Occasionally ER or PR are present in the cytoplasm instead of the nucleus. This is not interpreted as a positive result. Plasma cells have large amounts of cytoplasmic immunoglobulins that can crossreact with many antibodies. • In rare cases, immunoreactivity in an unusual location is of diagnostic importance:

• TTF-1: Cytoplasmic (instead of nuclear) positivity in hepatocellular carcinomas. • Ki-67 (MIB1): Cytoplasmic and membrane (instead of nuclear) positivity in trabecular hyalinizing adenomas of the thyroid and sclerosing hemangiomas of the lung.

• Beta-catenin: Nuclear (instead of cytoplasmic) positivity in solid pseudopapillary tumors of the pancreas and pancreatoblastomas. Both nuclear and cytoplasmic positivity is seen in the majority of colon carcinomas. Nuclear positivity is present in about 20% of endometrioid endometrial carcinomas and 70% of cases of desmoid-type fibromatosis. • ALK: The pattern of immunoreactivity correlates with the different types of chromosomal translocations in anaplastic large cell lymphomas. • NPM (nucleophosmin) shuttles between the cytoplasm and nucleus. NPM1 mutations occur in about 30% of adult AML and cause aberrant cytoplasmic expression of NPM ("NPMc+ AML"). These cases have a specific gene expression profile and distinctive clinical and prognostic features.

Examples of the normal location of antigens are shown in Figure 7 -1.

Identification of Immunoreactive Cells. Immunoreactivity of tumor cells must be distinguished from immunoreactivity of normal entrapped cells (e.g., desmin [+] skeletal muscle cells infiltrated by tumor, S100 [+] Langerhans cells in tumors, smooth muscle alpha actin [+] blood vessels, etc.). Plasma cells have large amounts of cytoplasmic immunoglobulin and can react nonspecifically with many antibodies. pathologists. The significance of immunoreactivity varies with the type of lesion, the antibody, and the specific assay. Strong positivity in the majority of cells is easily interpreted as positive. As the number of positive cells decreases, and the intensity of immunoreactivity weakens, the lower threshold of a "positive" result becomes more difficult to determine.

Time. Alkaline phosphatase chromogens (red color) fade over time. DAB (a brown color) is more permanent. This is not a problem in evaluating current pathology specimens. However, if slides are reviewed after a period of time, some chromogens may fade and once positive results may appear to be negative.

The following tables include information from the literature as well as the personal experiences of the staff at Brigham and Women's Hospital. Because of the many differences in specific antibodies, laboratory assays, and criteria for considering a result "positive," results may vary for different institutions. The results have been divided into five categories for general markers and four categories for hematopathology markers. Note that "%" refers to the number of tumors reported to be positive, not the number of cells positive within a tumor (Table 7 -2). The actual markers used to evaluate a case will depend upon the differential diagnosis based on the H&E appearance. In some cases, an initial panel, which is often used for typical cases, has been suggested. Not all markers listed would be used for all cases and some markers are included to indicate when they would not be useful for distinguishing the tumors listed in the table. POSITIVE is defined as the presence of immunoreactive cells and NEGATIVE as the absence of immunoreactive cells. Unfortunately, "positive" has also been used in some studies to mean "absence of immunoreactivity" when this finding supports a diagnosis. For example, the absence of SMAD4 (DPC4) has been reported as a "positive" result for pancreatic carcinoma, as this marker is absent in the majority of these tumors. This method of reporting results becomes confusing as "positive" and "negative" are dependent on the expected diagnosis. It is preferable to report the findings (positive = immunoreactivity present; negative = immunoreactivity absent) and then interpret them as supporting, or not supporting, the diagnoses in the differential diagnosis.

The combination of these two cytokeratins have been found to be useful to divide carcinomas into four main groups (Ck7+/Ck20+, Ck7+/Ck20-, Ck7-/Ck20+, Ck7-/Ck20-).

In Tables 7-3 to 7-7, other commonly used antibodies have been included to show differences within each group. The most useful additional antibodies will depend on the specific differential diagnosis. [6] [7] [8] Small blue cell tumors See Table 7 -8.

See Table 7 -9. About 50% of nasopharyngeal carcinomas are positive for Ck7. Many cases in Asian and North African patients (less commonly in US patients) are associated with EBV. EBV can be demonstrated by in situ hybridization, PCR, or occasionally by immunohistochemistry. These carcinomas are also positive for broad-spectrum keratins (AE1/AE3 and PANK). c S100-positive dendritic cells are present. P63 may be positive in breast "basal-like" carcinomas, some spindle cell metaplastic carcinomas, squamous cell carcinomas, and some papillary carcinomas. These subtypes may also have less typical keratin subsets such as Ck14 (detected by 34bE12), Ck17 (detected by MNF116), or Ck5/ 6. b Most well-and moderately differentiated ductal carcinomas, and carcinomas of special type (except for medullary) will be positive for hormone receptors. Poorly differentiated carcinomas, metaplastic carcinomas, and medullary carcinomas are usually negative. Well-and moderately differentiated lobular carcinomas are almost always positive for ER, and usually positive for PR. Poorly differentiated lobular carcinomas may be negative for these markers.

c Non-mucinous bronchioloalveolar carcinomas have an immunophenotype similar to lung adenocarcinomas. Mucinous BALs are more likely to be CK20 positive (about 70% positive), CDX2 positive, MUC2 positive, and less likely to be TTF-1 positive (about 30% positive).

d Secretory meningiomas are frequently positive for CK7 and CEA, whereas other subtypes are usually negative for Ck7 and CEA. The majority of all types of meningiomas are positive for PR (incuding meningiomas in males).

e Mixed tumors (pleomorphic adenomas) occur most frequently in the salivary glands, but can also arise in soft tissues (myoepithelial tumors of soft tissue). These tumors have a similar immunophenotype with keratin (AE1/AE3 77%)

or PANK (68%) or EMA (63%) present in the majority of tumors and frequent expression of markers associated with myoepithelial cells (e.g., calponin, GFAP, SMA, S100, p63). However, p63 is seen less frequently (23%) as compared to salivary tumors (100%).

f Chromophobe renal cell carcinomas may be positive for WT-1. Other types are negative. Rare colon carcinomas are either CK7 positive or CK20 negative, but they are rarely CK7+ CK20-. Although the majority of colon carcinomas are positive for CK20, almost one third of colon carcinomas with microsatellite instability (MSI-H positive) are CK20 negative. (see Table 19 -24). Significant immunoreactivity is a membrane pattern in the majority of the cells.

g Some plasma cell lymphomas may be positive.

Ewing's (PNET), desmoplastic small round cell tumor, rhabdomyosarcoma, neuroblastoma, and medulloblastoma have characteristic cytogenetic changes (see Table 7 -47).

EM has some advantages over immunohistochemistry in the evaluation of childhood small round blue cell tumors. 9 Initial panel: Keratin, S100, LCA. Additional studies may be helpful depending on the histologic appearance and the results of the initial studies. Keratin positivity may be present in ~25% of epithelioid angiosarcomas.

i Cellular dermatofbroma may show focal desmin immunoreactivity and a few will be CD34 positive.

j Alveolar soft part sarcomas are characterized by a translocation that fuses the TFE3 transcription factor gene at Xp11 to a novel gene at 17q25 called ASPL. These sarcomas demonstrate nuclear immunoreactivity for TFE3 (as do rare pediatric renal tumors with the same translocation) and this immunoreactivity is not present in other tumors or normal tissues. The characteristic cytoplasmic crystals are composed of monocarboxylate transporter 1 (MCT1) and its chaperone CD147. However, these proteins are found in many other cell types and are not specific for this tumor.

k Keratin and EMA positivity are usually only focal in monophasic synovial sarcomas.

l Claudin-1 is positive in glandular areas of synovial sarcoma but less so in spindle cell areas. m The immunohistochemical pattern for epithelioid mesotheliomas is given in a separate table.

n WT-1 may be positive in a minor epithelioid component of sarcomatoid mesotheliomas, but is generally negative in the spindle cells.

o Secretory meningiomas are typically cytokeratin 7 positive (Ck20 negative) and also positive for CEA. Other subtypes are generally negative for keratin. However, malignant meningiomas may be positive for keratin.

p Squamous cell carcinomas with a spindle cell morphology are generally strongly positive for AE1/AE3 (less commonly for CAM5.2), EMA, and p63. Spindle cell carcinomas of the breast often express markers expressed by myoepithelial cells such as "basal keratins" (including cytokeratin 14, which is included in the group detected by PANK or MNF-116), smooth muscle alpha actin, and p63. Poorly differentiated carcinomas with a spindle cell morphology may only show focal positivity for keratins and EMA.

q Dermatofibromas may have weak peripheral positivity for CD34 which is distinguished from strong diffuse positivity in DFSP.

r Other sarcomas (e.g., ~60% of MPNST) can be positive for MDM2 or CDK4. These markers are helpful to distinguish between benign and malignant lipomatous tumors.

Pathologists frequently receive specimens with metastatic tumors. Often, the site of origin is known to the clinician, but this information is not provided to the pathologist. A good clinical history is frequently more successful for correct classification than a battery of studies.

The Ck7/Ck20 pattern is generally helpful to narrow down the potential site of origin of carcinomas (see Tables  7-3 to 7-7) . Additional studies can then be used to identify specific types of carcinoma.

The most important tumors to identify are those with specific therapeutic treatments for cure or palliation ( Lung carcinomas are also TTF-1 positive, but will be thyroglobulin negative.

Highly effective treatment for cure with radioactive iodine.

Thyroid -medullary carcinoma Calcitonin If familial, important for counseling other family members.

Palliative treatment with tumordirected radionucleotides.

Inhibin Inhibin is not specific, but a trophoblastic tumor is unlikely if it is negative.

Chemotherapy for possible cure.

a Trastuzumab (Herceptin) = a monoclonal antibody directed against the HER2/neu receptor. b Imatinib mesylate (STI571, Gleevec™, Glivec™) is a small molecule tyrosine kinase inhibitor used for CML, ALL (Ph+), and GIST. The KIT protein is encoded by the c-kit proto-oncogene and is a transmembrane receptor protein with tyrosine kinase activity. Mutated proteins may or may not respond to therapy with Imatinib. Mutations that render KIT independent of its ligand, SCF (stem cell factor), have been found in GIST, AML, germ cell tumors and systemic mastocytosis. Wild-type KIT and KIT with mutations in the juxtamembrane domain (the intracellular segment between the transmembrane and tyrosine kinase domains) are found in GISTs and are sensitive to imatinib. Other tumor types are associated with mutations in the enzymatic domain and the altered protein is generally not sensitive to imatinib. c Gefitinib (Iressa) = a tyrosine kinase inhibitor effective against a small subset of lung adenocarcinomas with specific activating mutations. Some carcinomas may express unusual keratin sub-types. If negative, try other keratin types (e.g., CAM5.2). The Ck7/Ck20 pattern may be helpful in determining the likely site of origin. Some non-carcinomas can have an epithelioid appearance and strongly express keratins (e.g., epithelioid angiosarcoma, epithelioid sarcoma, mesothelioma).

Melanoma S100 protein S100 is strongly positive in the vast majority of melanomas. Some carcinomas (especially breast) and sarcomas are also positive for S100 and additional markers may be required. HMB-45 and MART-1 are expressed by most epithelioid melanomas but may be focal or absent in non-epithelioid melanomas (e.g., spindle cell or desmoplastic melanomas).

Leukocyte common antigen (LCA)

Present in almost all non-Hodgkin's lymphomas. May be absent in 30% of anaplastic (Ki-1) large cell lymphomas. These lymphomas are keratin negative but may express EMA. These tumors will be positive for CD30 (Ki-1) and ALK.

Hormone receptors are routinely determined on all invasive breast carcinomas and DCIS. ER and PR are weak prognostic markers and are more useful to predict the likelihood of response to hormonal therapies. Many different methods are currently used to report the results of IHC studies for ER and PR. One method that has been used in multiple studies is the Allred score method (Table 7 -12) .

Patients with carcinomas that scored 3 (<1% of cells with intermediate intensity or 1% to 10% of cells with weak intensity) or above had improved disease-free survival when treated with endocrine therapy. 11 Patients with carcinomas with a total score of 2 (<1% weakly positive cells) had a survival rate similar to ER-negative carcinomas (total score of 0). About 80% of DCIS cases are positive for ER using the same method of scoring. Women with ER-positive DCIS were shown to experience fewer local recurrences, contralateral recurrences, and distant recurrences when treated with tamoxifen (NSABP B24 study).

With optimization of IHC using newer antigen retrieval methods, 99.2% of carcinomas will have scores of 0, 7, or 8. 12 Therefore, many laboratories report results as positive or negative. The value of further subdividing cases by percent positive cells, H-score, or image analysis for either prognosis or to predict response to tamoxifen has not been demonstrated. Intensity of immunoreactivity is difficult to evaluate as most cases show strong reactivity with optimal assay methods and most carcinomas show considerable variability in intensity.

A possible method for reporting results is shown in Table 7 -13. The same system can be used for reporting progesterone receptor results. The use of both ER and PR may be helpful for determining the likelihood of response to tamoxifen, as has been shown with data using the biochemical assay (Table 7- 14) . Presumably, the presence of the ER-regulated gene product PR is more predictive of an intact ER regulatory pathway.

Recent national guidelines for reporting ER and PR have been released and should be consulted for additional information about performing and interpreting these studies (Hammond ME, et False negative results, and to a lesser extent, false positive results, can also be problems. False negative results may be due to a large number of causes including:

• Low sensitivity of the assay • Errors in performing the assay • Delayed fixation of tissue • Over-or underfixation of tissue • Overheating of tissue (e.g., with cautery during surgery) • Decalcification of tissue Most cases of false negativity can be suspected, as the normal breast tissue will also be negative. In such cases, the assay should be repeated on the same block, a different block from the same case, or blocks from another case, if available. If the normal tissue remains negative, the possibility of loss of antigenicity in the tissue can be mentioned in the report.

False positive results are quite unusual, as the antibody should not crossreact with other antigens.

• Entrapped normal ducts or lobules misinterpreted as carcinoma -this can be a difficult issue for DCIS as some ducts or lobules may be only partially involved by DCIS. • Control placed on same slide misinterpreted as the carcinoma • Sclerosing adenosis or myofibroblastoma (or other benign lesions) misinterpreted as invasive carcinoma

The HER2/neu immunoreactivity scoring system in Table  7 -15 was recommended by an expert panel. 13 Other panel suggestions:

• If cytoplasmic positivity obscures the membrane pattern, repeat the assay or perform FISH. • If normal ducts and lobules show definitive positivity, repeat the assay. • In cases of invasive carcinoma, only the areas of invasion should be scored. In some cases the associated DCIS can show stronger immunoreactivity. • Fixation must be in neutral buffered formalin and should, ideally, be between 6 and 48 hours for excisions, and at least 1 hour for needle biopsies. However, any effect from longer fixation has not been shown. • Unstained slides should not be used if prepared >6 weeks earlier. Loss of antigenicity has been shown for other antigens, but not specifically for HER2.

Only membrane immunoreactivity is scored. Marked cytoplasmic immunoreactivity may make interpretation difficult. FISH studies may be preferred for such cases ( In >90% of carcinomas with protein overexpression, the HER2/neu gene has been amplified. In 3% to 5% of cases, protein overexpression can occur due to other mechanisms. In <5% of cases, there may be gene amplification without protein overexpression. In general, there is a 20% to 40% response to Herceptin alone in patients with cancers showing gene amplification by FISH, and <5% of patients respond if the gene is not amplified. Therefore, FISH studies may be helpful for cases with 2+ positivity or difficult to interpret cases (e.g., with variable positivity or cytoplasmic positivity).

Well-and moderately differentiated lobular carcinomas are rarely positive (<5%). However, in some cases there may be edge enhancement of individual tumor cells that may be difficult to interpret. FISH studies may be helpful.

In rare cases, DCIS overexpresses HER2/neu but the accompanying invasive carcinoma does not. This is a source of potential false positive results for IHC or FISH.

Myoepithelial markers can be useful for the evaluation of breast lesions ( can show focal to diffuse positivity for myoepithelial markers. S100 protein and cytokeratins (e.g., 34βE12) are not recommended for identifying myoepithelial cells, as fewer myoepithelial cells are positive and luminal cells can also be positive. p63 is a good general marker for myoepithelial cells and is particularly helpful in cases with prominent myofibroblasts (e.g., sclerosing lesions) or with blood vessels closely apposed to tumor cells (e.g., papillary fronds in papillary DCIS). In some cases, SMA may be positive in more myoepithelial cells than p63. 

Most cases of Paget disease of the nipple are associated with DCIS deeper in the breast that involves the lactiferous sinuses, and about half will also have areas of invasion. Rare cases may be difficult to interpret due to the absence of associated disease in the breast or if the initial biopsy is shallow. In some cases, Paget cells may take up melanin and may be difficult to distinguish from melanoma. Toker cells are present in nipple epidermis in 40% to 80% of nipples and are Cam5.2 and Ck 7 positive but are negative for HER2/neu. Paget disease of the vulva and perianal region has a similar distribution (i.e., tumor cells are present between an intact basement membrane and an overlying normal epidermal layer) but the tumor cells have different origins. Initial panel: Cam5.2 (or Ck7), HER2, and S100 with additional antibody studies based on these findings, if necessary. 

See Metastatic lobular carcinoma of the breast can morphologically resemble primary signet ring cell gastric carcinomas. Both typically lack e-cadherin expression. In addition, lobular breast carcinomas can be clinically occult or can present as distant metastases many years after the initial presentation. Signet ring cells associated with breast carcinoma more commonly have a central mucin vacuole with a targetoid appearance (cell A). Gastric signet ring cells usually have many small vacuoles giving the cytoplasm a foamy appearance (cell B). These criteria are not reliable in distinguishing these two carcinomas. However, the presence of the first type of signet ring cell in a biopsy from the gastrointestinal tract should raise the possibility of metastatic breast carcinoma. 14 The majority of lobular breast carcinomas will be ER positive, and this is a reliable marker to exclude gastric carcinoma. In the minority of ERnegative cases, PR, GCDFP, MUC1, CDX2, and Hep Par may be helpful markers. 15

See Table 7 -21.

See CK 14 and CK 17) . These keratins may be detected best with MNF-116 (= PANK; includes CK 17) or 34betaE12 (includes CK 14) or antibodies specific for these keratins. Some epithelioid myofibroblastomas can closely resemble invasive lobular carcinoma. In these cases, the carcinoma will be strongly positive for typical keratins and also positive for ER and PR. * The lack of SMAD4 is found in about half of pancreatic carcinomas and is highly suggestive of this primary site. In some published tables a "positive" result is the absence of positivity. In this table, "positive" signifies that the carcinoma shows immunoreactivity for SMAD4.

See Table 7 -27.

See Table 7 -28.

See Table 7 -29. 

See Table 7 -30. FISH for 12p can be used to identify germ cell tumors and their metastases. D2-40 (podoplanin) is strongly expressed in seminomas and ITGCN. It is also expressed in lymphatic endothelium, epithelioid mesotheliomas, and hemangioblastomas.

See Table 7 -31. Table 7 -46). c Antibody A103 is positive in adrenal cortical carcinomas. Another antibody to the same antigen, M2-7C10 is not positive in adrenal cortical carcinomas.

d Clear cell renal cell carcinoma metastatic to the adrenal can sometimes be confused with an adrenal cortical tumor (thus, the older term for clear cell carcinoma of "hypernephroma"). RCC has clear cytoplasm (compared to the bubbly cytoplasm of the adrenal cortex) and blood lakes are typically present. Glycogen is present in RCC and absent in adrenal lesions (demonstrated by PAS with and without diastase). Cytokeratin and EMA are useful IHC markers. Diff = diffuse positivity; mem = positivity located on membrane.

Renal cell carcinoma subtypes have typical cytogenetic abnormalities (see Table 7 -47). CD117 (c-kit) has been reported to be positive in almost all papillary renal cell carcinomas (cytoplasmic) and chromophobe carcinomas (membrane) but is not present in clear cell carcinomas. Mutations in c-kit were only found in papillary carcinomas.

See Table 7 -32.

See Table 7 -33. Antibody cocktails: These antibodies can be combined to facilitate the evaluation of small lesions: 34βE12 + p63 = labels a greater number of basal cells than either marker alone. AMACR + p63 and/or 34β12 = facilitates the identification of small foci of invasive carcinoma.

See Table 7 -34. TTF-1 is seen in the cytoplasm (unlike the nuclear pattern seen in lung and thyroid carcinomas)

c Carcinoids arising at sites other than lung are very unlikely to be positive for TTF-1. Lung carcinoids may be positive and are more likely to express Ck7.

d Mucin histochemical stains can also be used. HCCs will be negative and 75% to 100% of cholangiocarcinomas will be positive. Cyt = cytoplasmic immunoreactivity; nuc = nuclear immunoreactivity. Sinusoids of HCC show diffuse CD34 positivity in 80% to 90% of cases, but this is not seen in normal liver. CD34 positivity can also be seen in focal nodular hyperplasia. Metastatic carcinomas can show diffuse positivity in 20% of cases, but the positive endothelial cells are present throughout the tumor and the cells do not surround nests of tumor cells, as is seen in HCC.

Reticulin stains can be helpful in the evaluation of fine needle aspirates or core needle biopsies of liver lesions. HCC has an abnormal pattern of absent, decreased, or expanded trabecula, whereas benign lesions will show a normal trabecular pattern. Metastatic carcinomas can usually be distinguished from HCC by frequent expression of Ck7, only rare expression of HepPar1, the absence of a bile canicular pattern for CEAp and CD10, and the absence of cytoplasmic positivity for TTF-1. Metastatic carcinomas to the liver often cannot be reliably distinguished from cholangiocarcinomas by histologic appearance or immunohistochemical pattern, with the exception of colorectal carcinomas. If the patient has a known primary carcinoma, it is most helpful to compare the two tumors.

See Table 7 -35. Tumors with Hurthle cell changes may be negative for HBME. b Hurthle cells (both benign and neoplastic) are positive for S100 (nuclear and cytoplasmic). c Spindle cells may be positive for vimentin.

d Anaplastic thyroid carcinomas are frequently negative for TTF-1, thyroglobulin, and Ck20 but positive for p53 and Cyclin D1. e p27 is low in parathyroid carcinomas.

Thyroid adenomas, follicular carcinomas, papillary carcinomas, and medullary carcinomas are Ck7+ and Ck20-. Variable immunoreactivity has been reported for Ck7 in anaplastic carcinomas. Metastatic carcinomas to the thyroid will be negative for thyroglobulin, TTF-1 (except for lung carcinomas), and calcitonin.

DDIT3 and ARG2 are new markers that may prove helpful for distinguishing follicular carcinoma (~ 70-80% positive) from adenoma (90% negative)

See Table 7 -36. Initial panel: AE1/AE3, calretinin, WT-1 (clone 6F-H2), CEA, Leu-M1, and TTF-1 with additional studies ordered in difficult cases.

Other antibodies generally reported as negative in epithelial mesotheliomas and positive in lung adenocarcinomas include the following: MOC-1, B72.3, Ber-EP4, and BG-8. Cytokeratins 5/6 are reported to be positive in mesotheliomas and negative in lung carcinomas. However, in our experience, these markers have proven less useful than the ones listed earlier. The use of EMA is controversial. Strong membrane positivity is characteristic of epithelial mesothelioma, whereas cytoplasmic positivity is characteristic of adenocarcinomas.

Less is known about the immunophenotype of pure sarcomatoid mesotheliomas. The spindle cells are positive for cytokeratin, but are less frequently positive for the other markers as compared to the epithelioid cells. Tumors that can, on occasion, resemble mesotheliomas are generally negative for cytokeratins, with the notable exceptions of some cases of angiosarcoma, epithelioid hemangioendothelioma, synovial sarcoma, epithelioid sarcoma, and leiomyosarcoma (see Table 7 -9). 16 Deletions of 3p, highly variable changes a Keratin immunoreactivity is accentuated around the nucleus and is present in the cytoplasm, without a prominant membrane accentuation. b Keratin immunoreactivity is diffusely present in the cytoplasm with membrane accentuation in some cells. c WT-1 immunoreactivity is nuclear. d Metastatic adenocarcinomas are generally negative for WT-1 except for ovarian serous carcinomas and some renal carcinomas (see Table 7 -5) . e Most metastastic adenocarcinomas will be positive for CEA, but there are some exceptions (see Table 7 -5).

Tissue should be obtained for EM and cytogenetics, if possible. 

See Tables 7-37 and 7-38. 

Amyloidosis (Greek for amylon = starch plus eidos = resemblance) is seen in many different clinical settings and is associated with many diseases. Pathologists can narrow down the differential diagnosis considerably to help guide clinical decision making. Finding an amyloid deposit in any tissue is similar to finding metastatic carcinoma in a lymph node -in both settings clinical information (e.g., history, physical examination, radiology studies, results of laboratory tests) is essential in arriving at the correct interpretation. A little immunohistochemistry and a lot of clinical judgment by the pathologist can help establish the cause with a greater degree of certainty. 17 Finding and characterizing amyloid deposits: 1. Examine the H&E slide for noncellular material in the correct location for the suspected disease (Table 7 -42). a Light chains are detected best by immunofluorescence (IF) on unfixed frozen tissue. IF and IHC can be performed on paraffin sections, but with less specificity. Only 50% of cases of light chain disease amyloid will be positive because the amyloid protein is often derived from the variable domain whereas antibodies detect the common domain. b Serum amyloid protein A can only be detected by IF on unfixed frozen tissue.

Amyloid deposits will be orange-pink on Congo Red stains or sea-foam green on Sulfated Alcian blue stains. Amyloid may be more apparent on these stains. HOW-EVER, beware of overcalling cases in which there is not a histologic correlate for amyloid in the stained tissue.

If there is background positivity in normal tissue due to overstaining, the slide cannot be interpreted. Positive controls must show appropriate specific positivity.

Congo red-positive amyloid should become an apple green color when viewed under polarized light. This may require the high-quality polarizers that are built in to the microscope. Lower quality polarizers (i.e., the cut squares of polarizing material) may not be adequate. Collagen (silver when H&E is polarized) and fibrin (does not polarize) may mimic amyloid. 4. The amyloid deposits can be further characterized using immunohistochemistry or immunofluorescence (see Table 7 -42) based on the clinical information, the organ or structures involved, and the distribution of amyloid deposits in the tissue. Amyloid can also be identified using EM (non-branching fibrils, 7.5 to 10 nm width and up to 1 micron in length).

A firm diagnosis is not always possible. The final diagnosis should be based on a combination of histologic, immunohistochemical, and clinical data.

See Tables 7-43 and 7 -44.

The results of studies are incorporated into the surgical pathology report. The following information is included:

1. The type of tissue studied: formalin-fixed (or other fixatives) tissue, cryostat sections, cytology preparations, etc. 2. The type of immunoagents used, being as specific as possible. For example, do not just list keratin but specify the type of keratin (e.g., AE1/AE3).

The results of the studies in great enough detail to allow interpretation. For example the type of cell that is immunoreactive (e.g., tumor vs. nontumor), intensity of immunoreactivity (e.g., weak, strong) and/or the number of cells immunoreactive (e.g., focal vs. diffuse). 4. Integration of the results into the final diagnosis specifying whether they confirm or support a diagnosis, make one diagnosis more likely than others, or exclude one or more diagnoses.

Text continues on page 157. Notes: NAME: The most common name used to refer to the marker. The name may refer to the antigen, a CD number, or a specific antibody raised to the antigen. In some cases more than one name is commonly used. Underlined antibodies appear in the tables. Most CD numbers correspond to a specific gene product. However, some correspond to antigens from post-translational modifications. 

Indications for EM Studies.

• Diagnostic renal biopsies for glomerular disease • Adenocarcinoma versus mesothelioma (see Table 7 -36) • Difficult to classify tumors (Tables 7-45 and 7-46) • Nerve (e.g., toxic or drug-induced neuropathy) and muscle biopsies (e.g., inclusion body or nemaline myopathy) • Bullous skin diseases (e.g., epidermolysis bullosa) • Ciliary dysmorphology (primary ciliary dyskinesia or Kartagener syndrome) • Endomyocardial biopsies (e.g., adriamycin toxicity, amyloid, nemaline myopathy) • Liver biopsies for microvesicular fat in acute fatty liver of pregnancy • Small bowel biopsies to look for pathogens (e.g., Whipple disease) • Congenital, inherited, and metabolic diseases (e.g., ceroid lipofuscinoses) • Prion diseases Method. Ultrastructural details of tissues are lost rapidly. Therefore, fresh tissue must be fixed rapidly and well for EM. Tissues are usually fixed in special fixatives for EM to preserve lipids and glycogen (e.g., 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4).

Place a small fragment of tissue in a drop of fixative on a cutting surface. 2. Cut the tissue into multiple tiny fragments, each no greater than 0.1 cm in any dimension. 3. Place the tissue into the vial of fixative. Shake the vial to make sure all the tissue fragments are covered by fixative.

Note: If tissue from a small biopsy is found to be nondiagnostic on H&E, any tissue saved for EM should be retrieved for examination by light microscopy.

A separate electron microscopy report is usually issued. The results should be incorporated into the final diagnosis. 19

Rhomboid, rod-shaped, or spiculated crystals in a regular latttice pattern.

The characteristic cytoplasmic crystals are composed of monocarboxylate transporter 1 (MCT1) and its chaperone CD147. These proteins are found in many other cell types and are not specific for this tumor. Cytogenetics: t(X;17) creates a ASPL-TFE3 fusion protein. IHC: TFE3 positive (as well as rare pediatric renal tumors with the same translocation). Nuclear immunoreactivity is not present in other tumors or normal tissues. Histo: The crystals are PAS with diastase positive. Other tumors can also be keratin positive and have desmosomes, filaments, and cytokeratin (mesothelioma, meningioma, synovial sarcoma, and epithelioid sarcoma)

Melanosomes in various stages of development -indicative of a melaninforming cell type. Abnormal pleomorphic melanosomes may be present in melanomas. Desmoplastic melanomas lack melanosomes.

IHC: S100, HMB45, MART-1 HMB45 and MART-1 may be absent in non-epithelioid melanomas. The HMB-45 epitope (gp100) is present in immature melanosomes or premelanosomes, but is not specific to these structures.

Melanosomes are also seen in clear cell sarcoma, pigmented schwannomas, PEComa, and other rare tumors. Mature forms can be taken up by melanophages, keratinocytes, and carcinomas.

No specific features are present. The cells lack cellular junctions and there is a paucity of cytoplasmic organelles.

IHC: LCA LCA may be absent in 30% of anaplastic (ALK1) lymphomas. These tumors can be EMA (+) but are keratin (-).

Some types have specific diagnostic features of cell type (e.g., neural, smooth muscle, striated muscle). No well-developed desmosomes.

IHC: May be helpful for identifying specific types.

Keratin negative except for synovial sarcoma and epithelioid sarcoma (or rarely in other types).

Amyloid Non-branching fibrils, 7.5 to 10 nm in width and up to 1 micron in length.

May be present associated with plasma cell tumors, medullary carcinoma of the thyroid, Alzheimer's disease, or as an isolated finding (primary amyloidosis). IHC: Can be used to identify specific types of amyloid (e.g., lambda or kappa chains, beta2 microglobulin, calcitonin, tau) Bronchioloalveolar carcinoma of the lung (BAL) Lamellar (surfactant) "myelin-like" granules in the supranuclear cytoplasm (typical of Type II pneumocytes). Clara-like electron-dense granules in supranuclear cytoplasm. Intranuclear inclusions comprised of parallel microtubular arrays. These features can also be seen in metastatic BAL.

Cytogenetics: These carcinomas are less likely to be associated with smoking and have fewer cytogenetic changes. Bronchioloalveolar carcinomas or adenocarcinomas with features of BAL are more likely to respond to Iressa (38%) as compared to other lung carcinomas (14%) due to specific mutations in EGFR. Mucinous BAL has intranuclear inclusions but generally lacks the other EM features. Cytogenetics: t(12;22) EWS;ATF1 fusion protein IHC: S100, HMB45

Dense core granules Dense core granules (vesicle bound by a single membrane with a dense center -60 to 300 nm) -cytoplasmic organelles involved in regulated exocytosis of cell products. Examples: Pancreatic beta cells (insulin): angular crystalline inclusions Pheochromocytoma (epinephrine and norepinephrine): Large, pleomorphic, often clear or only partially filled Carcinoid:

• Foregut -small, round • Midgut -larger, pleomorphic • Hindgut -mixed Found in tumors of neuronal or neuroendocrine origin. Vesicles are comprised of granins (predominantly chromogranin A, chromogranin B, and secretogranin II) and various peptide hormones and transmitters, ATP, and biogenic amines IHC: Chromogranin A (most specific). Specific products of tumors can also be detected. Note: Prostate cancers and breast cancers can also show strong chromogranin positivity and can be mistaken for neuroendocrine tumors, particularly at metastatic sites. Weibel-Palade bodies are frequently absent in tumors arising from endothelial cells (e.g., angiosarcomas). IHC markers are more sensitive to detect endothelial derivation. The membranes are formed by P-selectin and the tubules contain FVIII. IHC: Vascular markers (CD34, CD31, FVIII)

Homogeneous cell population characterized by the lack of specialized features, large pools of glycogen, no organelles, no extracellular matrix, variable numbers of neurosecretory granules and cell processes.

Cytogenetics: t(11;22) EWS;FLI1 fusion protein (and other less common variants) IHC: CD99. FLI1 is also present, but is less specific. Histo: PAS +/− diastase can detect glycogen, but is not currently used for diagnosis.

Numerous lysosomes (filled with tubular, vesicular, and amorphous material), phagosomes, and granules (correlating with the "granular" cytoplasm), reduplicated basal lamina surrounding groups of cells.

IHC: S100, inhibin, CD68, calretinin

Birbeck granules (rod-or tennis racket-shaped) structures of variable length with a central periodically striated lamella.

May serve as a reservoir for Langerin (a transmembrane type II Ca2+-dependent lectin) and CD1a in the endosomal recycling compartment. IHC: CD1a, Langerin, S100

Lameller or scroll-like membrane pattern, granules of variable size. Cytogenetics: Characteristic changes in alveolar and embryonal types IHC: Muscle markers (HHF-35, desmin, myf4)

Basal lamina prominent, often reduplicated. Luse bodies (long spacing collagen, extracellular), myelin figures, long cell processes wrapping around collagen, may rarely have melanosomes (melanotic schwannoma)

Cytogenetics: Deletion of 2q (NF2 inactivation) IHC: S100 

Frozen tissue is useful for staining (some antibodies only detect antigens in frozen tissue), enzyme studies (muscle biopsies), and to save tissue for DNA or RNA studies.

All specimens with a question of a lymphoproliferative disorder, sarcomas, unusual tumors, muscle biopsies.

Methods. Small (approximately 0.5 × 0.5 × 0.3 cm 3 )

portions of tissue are placed in a clean specimen container moistened with a small amount of normal saline until they can be frozen. Specimens should be snap frozen using liquid nitrogen or dry ice and stored at -20°C.

The results of studies on frozen tissue are usually incorporated into the surgical pathology report.

Like immunoperoxidase studies, immunofluorescence (IF) detects antigens in tissues. However, because amplification of the signal is not used, it is better suited for precise localization of antigen/antibody complexes in tissues or for determining the deposition pattern of immune complexes (e.g., linear vs. granular). Thus, it is most useful for the investigation of diseases related to immune complex deposition such as glomerular diseases and bullous diseases of the skin. Tissue for IF may be snap frozen (see instructions earlier) or stored in special fixatives for IF. If the specimen is not frozen, special care must be taken to ensure that the biopsy is kept moist in a sealed container.

• Direct IF: Uses antibodies to detect antigens in the patient's tissues. • Indirect IF: Uses control tissues to detect antibodies (e.g., anti-BM) in the patient's serum.

Indications. Some skin biopsies (e.g., lupus, pemphigus, pemphigoid, and dermatitis herpetiformis), all diagnostic nontransplant renal biopsies, some transplant renal biopsies, identification of amyloid in cardiac biopsies, and the evaluation of vasculitis in nerve biopsies.

Method. Tissue must be submitted fresh.

Results. The results of the examination are usually incorporated into the surgical pathology report.

• SLE (lupus band test): linear or granular staining along dermal epidermal junction for multiple immunoreactants in about 80% of cases (most commonly IgG and less often IgM or C3). The specificity increases with the number of positive immunoreactants. Uninvolved sun-exposed skin shows positivity in most patients with active systemic lupus. Uninvolved skin in patients with discoid lupus is usually negative for this test. • Herpes gestationis: perilesional skin shows linear BM zone C3 and sometimes IgG. • Dermatitis herpetiformis: granular IgA at tips of dermal papillae of uninvolved skin. • Pemphigus: IgG and C3 between epidermal cells creating a net-like pattern. In pemphigus vulgaris, a split just above the basal cell layer creates a "tombstone" appearance to the row of basal cells at the base of the vesicle.

In pemphigus foliaceus and related disorders, the split occurs near the granular cell layer. • Pemphigoid: Ig and C3 along basement membrane but not between cells. Indirect IF reveals an anti-BM antibody.

Molecular genetic pathology is the newest subspeciality in pathology with board certification. Molecular diagnostics incorporates many types of techniques for the investigation of genetic alterations in cells and viruses (e.g., Southern blotting, PCR analysis, FISH). It has applications in three main areas:

Inherited diseases: • Identification of inherited diseases (e.g., cystic fibrosis, hemochromatosis, Factor V Leiden, Prothrombin 20210A, Fragile X syndrome). • Identification of genes conferring susceptibility to diseases (e.g., microsatellite instability [MSI], BRCA1 and 2)

Infectious diseases: • Detection of organisms • Identification of specific organisms • Quantitation of viral infection (e.g., HIV viral load)

• Identification of specific genetic alterations associated with tumors • Identification of gene mutations associated with susceptibility to treatment (e.g., EGFR mutations in lung cancer, c-kit mutations in GIST) • Identification of clonality in hematolymphoid proliferations • Detection of minimal residual disease after treat ment These studies are especially helpful for hematolymphoid proliferations that are difficult to classify because of the frequent and characteristic rearrangements that occur in many of these disorders. Unlike cytogenetics, the cells need not be viable. However, it is preferable that the nucleic acids are relatively intact. Southern blot and RNA-based PCR (RT-PCR) assays are best performed on fresh or frozen tissues. Formalin-fixed, paraffin-embedded tissue is amenable to DNAbased PCR assays. Some fixatives (e.g., Bouin's) cause extensive breakage of DNA and may preclude genetic analysis of the tissue.

• B-cell proliferations -clonal rearrangements of the immunoglobulin heavy and light chain genes; specific translocations • T-cell proliferations -rearrangements of the γ and β Method of Submitting Tissue. Fresh or frozen tissue (e.g., snap frozen tissue) as well as fluids may be used. Cytologic preparations can be used for FISH. Paraffin blocks can also be used.

The results are usually either reported separately or incorporated into the surgical pathology report.

Cytogenetic studies have been demonstrated to be useful in several areas important to pathology:

• Tumor classification: Particularly sarcomas (e.g., Ewing's sarcoma and synovial sarcoma), lymphomas, leukemias, kidney tumors, brain tumors, and other unusual tumors. • Benign vs. malignant lesions:

• Reactive mesothelial cells vs. mesothelioma • Lipoma vs. liposarcoma • Prognosis: Neuroblastoma, oligodendroglioma, multiple myeloma, chronic lymphocytic leukemia. • Treatment: Amplification of HER2/neu to predict response to Herceptin. • Research: Translocations are common to many tumors and usually identify genes important to the pathogenesis of the tumor Cells may be cultured to perform complete karyotype analysis or tissues can be analyzed for specific chromosomal alterations by fluorescence in situ hybridization (FISH) .

FISH studies can be performed on cultured cells, cytology specimens, touch preparations, and paraffin-embedded tissues.

• For karyotype analysis: Soft tissue tumors, mesotheliomas (tissue or pleural fluid), unusual tumors, poorly differentiated tumors, all subcutaneous lipomas >10 cm or of unusual gross appearance, all deep-seated lipomas (subfascial, intramuscular, intraabdominal, retroperitoneal, clinically apparent cord tumors), unusual uterine masses. • For FISH: Oligodendroglioma, neuroblastoma.

Method for Submitting Tissue. Tissue for karyotyping must be fresh, viable, and relatively sterile. However, tissue may be submitted even if it has not been handled under strictly sterile conditions (contamination is not usually a problem). If specimens are to be held overnight, the tissue should be minced (into 0.1 cm cubes) in a sterile specimen container, covered with culture medium, and held overnight in the refrigerator. Fluids may also be submitted for analysis (especially pleural effusions with a suspicion of mesothelioma).

The results of the cytogenetic analysis should be incorporated into the final diagnosis or reported separately (Tables 7-47 and .

Text continues on page 178. 

Isochromosome 17q >25%

1p deletion 25% (CD117) is encoded by the c-kit proto-oncogene and is a transmembrane receptor protein with tyrosine kinase activity. Mutations may render KIT independent of its ligand, SCF (stem cell factor). Mutated proteins may or may not respond to therapy with Imatinib. Wild-type KIT and KIT with mutations in the juxtamembrane domain (the intracellular segment between the transmembrane and tyrosine kinase domains) are found in GISTs and are sensitive to imatinib. Other tumor types are associated with mutations in the enzymatic domain and the altered protein is generally not sensitive to imatinib. Overexpression of the protein is detected by IHC. d Gefitinib (Iressa) = a tyrosine kinase inhibitor effective against a small subset of lung adenocarcinomas with specific activating mutations in EGFR. IHC for EGFR is not helpful for identifying carcinomas likely to respond to treatment. For additional information on specific genes, see Online Mendelian Inheritance in Man (OMIM; www.ncbi.nlm.nih.gov). a Imatinib mesylate (STI571, Gleevec, Glivec) is a small molecule tyrosine kinase inhibitor that may be used for the treatment of tumors overexpressing tyrosine kinases: Bcr-Abl tyrosine kinase: CML, ALL (Ph+) KIT tyrosine kinase: GIST, systemic mastocytosis, some types of AML PDGFR kinase: CMML, chronic eosinophilic leukemia, rare cases of GIST The KIT protein is encoded by the c-kit proto-oncogene and is a transmembrane receptor protein with tyrosine kinase activity. Mutations may render KIT independent of its ligand, SCF (stem cell factor). Mutated proteins may or may not respond to therapy with Imatinib. Wild-type KIT and KIT with mutations in the juxtamembrane domain (the intracellular segment between the transmembrane and tyrosine kinase domains) are found in GIST's and are sensitive to imatinib. Other tumor types are associated with mutations in the enzymatic domain and the altered protein is generally not sensitive to imatinib. For additional information on specific genes, see Online Mendelian Inheritance in Man (OMIM; www.ncbi.nlm.nih.gov).

The following features are suggestive of hereditary susceptibility to cancer:

• Two or more close relatives on the same side of the family with cancer • Evidence of autosomal dominant transmission • Early development of cancer in the patient and relatives (in general, <50 years of age) • Multiple primary cancers • Multiple types of cancers • Unusual pathologic features of tumors (Table 7-49) • A constellation of tumors suggestive of a specific syndrome (Table 7-50) Pathologists can aid in the detection of hereditary carcinomas by being aware of the types and pathologic characteristics of carcinomas associated with these syndromes. Patients with germline mutations are important to identify in order to:

• Screen patients for other common tumors or other components of the disease • Consider prophylactic surgery or preventive interventions • Offer screening to family members at risk and genetic counseling

Text continues on page 184. Colon carcinomas are more likely (overall, 66%) to be on the right side, poorly differentiated ("medullary"), mucinous, signet ring, or undifferentiated, with a prominent lymphocytic infiltrate. IHC can be used to detect the absence of MSH2 (usually due to germline mutations) and MLH1 (can be due to germline mutations, epigenetic changes (methylation), or less commonly, somatic mutations) in many patients. IHC is 92% sensitive for MSI and 100% specific. MSI testing is also used. 

Although the sporadic forms of cancers, in general, far outnumber cases associated with germline mutations, in some cases the appearance or site of a carcinoma is highly suggestive of a known syndrome and further investigation may be warranted.

Flow cytometers analyze populations of thousands of dissaggregated cells as they pass by stationary detectors. Cell size and cytoplasmic granularity can be measured as well as DNA content and the presence or absence of immunohistochemical markers added to the cell suspension. Newer techniques can analyze three or more features simultaneously to divide cells into unique populations. DNA content can be used to determine the number of cells in S-phase (a measure of proliferation -S-phase fraction). Because cells are not visualized by this technique, one must be sure to submit only lesional tissue.

• Hydatidiform moles -complete (diploid), partial (triploid) • Some carcinomas -DNA ploidy and S-phase have been reported to be of prognostic significance for some carcinomas (e.g., colon, breast, and prostate) but is not routinely performed at all institutions or used by all oncologists.

Indications for Cell Surface Marker Analysis.

Method for Submitting Tissue. Single cell suspensions are necessary for analysis. For fresh tissues, cells must be viable. Fresh tissue (approximately 0.3 to 0.5 cm 3 ) is placed in a specimen container and kept moist with HBSS. Tissues can be held overnight in a refrigerator. Formalin-fixed paraffin-embedded sections may also be used for DNA ploidy analysis by the Hedley method, although the results are not as satisfactory due to nuclear fragmentation.

The results are usually incorporated into the final surgical pathology report.

Cytologic preparations of surgical specimens often add additional information.

• Intraoperative diagnosis: Touch preps or smears are especially valuable for: • Infectious cases (to avoid contamination of the cryostat and aerosolization of infectious agents) • Neuropathology cases -for diagnosis and for the performance of cytogenetic (FISH) analysis.

• Tumors (for excellent cytologic detail, especially lymphomas and papillary carcinomas of the thyroid) • Special stains: Stains for microorganisms can be performed the same day on cytologic smears of specimens from critically ill patients. Do not submit air-dried smears of infectious cases for staining as the unfixed material may constitute a hazard to laboratory personnel. Fat is dissolved during routine processing, but can be demonstrated with fat stains on air dried slides. • Genetic studies (FISH): In touch preparations nuclei are intact, unlike tissue sections in which only partial nuclei are present. This feature makes these preparations superior for techniques such as FISH and image analysis.

Comparing cytology preparations and the corresponding surgical specimen is always a useful exercise in learning the comparative morphology of these techniques.

Specimen radiographs are often preferable over patient radiographs:

• A permanent record of the radiograph can be kept with the case. • A radiograph of the specimen may reveal more details of the underlying process (e.g., fewer structures may be present to complicate the appearance). • There may have been a significant time interval between the patient radiograph and the surgical excision. • The radiograph will often indicate important sites to examine histologically (tumor invasion into a rib or microcalcifications in a breast biopsy). • The specimen radiograph can confirm that the clinical lesion was removed.

• Tumors of bone and cartilage • Tumors invading into bone • Avascular necrosis • All bioprosthetic heart valves (to document the degree of calcification) • Breast biopsies or mastectomies performed for mammographic lesions that cannot be located grossly. Paraffin blocks of breast tissue can be radiographed if microcalcifications were seen by specimen radiography but not in histologic sections and were not identified prior to processing. Clips placed after core needle biopsy are also easily identified.

Calcifications can dissolve in formalin over several days. If the demonstration of calcifications is important (e.g., mammographically detected calcifications) it is preferable to process the tissue within 1 to 2 days. If processing is to be delayed, the tissue can be stored in ethanol.

Radiographic equipment is available in radiology departments and in some pathology departments. The specimen may be placed on a piece of wax paper (to keep the surfaces clean) lying on the film. Specimens can be radiographed after decalcification (not all calcium is removed) but best results are obtained on fresh undecalcified specimens. Lungs should not be inflated prior to radiography.

If the specimen is small, two exposures at different settings or at different angles may be useful. Lead sheets can be used to allow two exposures on one piece of film.

If the film is too dark (overexposed), the exposure is too high and a lower setting should be tried. If the film is too light (i.e., unexposed) the exposure is too low and a higher setting is indicated.

Special injection techniques with radiocontrast media are available for unusual specimens (e.g., a recipient lung with pulmonary hypertension, vascular ectasia of the bowel).

Octreotide and Sentinel Nodes. Labeled compounds are sometimes used to localize certain types of tumors (generally neuroendocrine) or sentinel lymph nodes. The patient is injected with the isotope prior to surgery and the surgeon uses a hand held probe to identify the labeled tissue. The amount of radioactivity in the tissue is small and generally does not pose a hazard to pathologists handling the tissue and does not need special disposal methods. However, each pathology department should consult with their radiation safety department to ensure appropriate handling of such tissues. In some cases, if a gross lesion is not present corresponding to the area of octreotide uptake, specimens can be imaged using a gamma camera.

Results. The radiographs are documented in the gross description and any information gained from the radiograph is incorporated into the surgical pathology report.

The pathology department is a unique resource for researchers who need human tissues. The pathologist plays a key role as patient advocate and diagnostician in order to provide appropriate human tissues for biologic research. Most hospitals have a policy that allows the release of tissue for research if it would otherwise be discarded. Therefore, tissue is never provided for research until all necessary tissue has been taken for diagnosis. Tissue from primary diagnostic breast biopsies and open lung biopsies without gross lesions must not be given away. It is in the best interest of the patient that a pathologist evaluate the specimen rather than have tissue given away by nonpathologists who are not aware of what is needed for diagnosis.

Indications. By request of researchers who have obtained permission from the hospital Institutional Review Board (IRB). Patients must provide specific consent. In some cases, all patient identification will need to be removed from the specimen.

The investigation of infectious disease by culture is complementary to its investigation by histologic sections (Tables 7-51 to 7-53). Can be performed on aspirates, swabs, fluids, or tissues. Requires surgical excision of tissues.

Cultures amplify the number of organisms present, allowing them to be recognized.

Organisms may be rare, or not seen in tissue sections.

The specific organism can be identified and tested for drug susceptibility.

Categories of organisms can be recognized but specific identification may not be possible.

Some organisms cannot be cultured. Many organisms can be identified that will not grow in culture or that require long culture times (e.g., TB).

It may be difficult to exclude contamination for a positive culture.

Morphologic evidence of an inflammatory response provides evidence for a clinical infection. The location of the infection may be of diagnostic importance (e.g., cellulitis vs. necrotizing fasciitis or superficial colonization of devitalized tissue vs. deep infections involving viable tissues). The use of special studies, such as PCR and other molecular assays, may be warranted if the tissue pattern of injury is indicative of a potential organism despite lack of culture evidence and/or lack of organisms seen on tissue sections. 

• Suspected infectious processes, either by clinical data or by frozen section • Suspected sarcoid to exclude an infectious process Method. Tissue is kept as sterile as possible. Suture removal kits are a convenient source of sterile scissors and forceps. Serially section the specimen to determine if there are focal lesions. Place representative sections in a sterile specimen container making sure to retain a duplicate piece of tissue for histology. Label with the patient's name and unit number, patient's physician, type of specimen, collection date, and time of collection (required for Joint commission accreditation).

The results are generally reported by the microbiology laboratory. Communication with the microbiology laboratory and staff is essential to correlate histologic results with microbiology results from the same specimen.

Reports. The results of culture of surgical specimens are usually reported in a separate report. 

AI, acute inflammation

ISH, in situ hybridization methods

MP, macrophage; MSS, silver stain (similar to GMS-Grocott-Gomori methenamine silver)

Pathologic Diagnosis of Fungal Infections

* Cowdry type B inclusions were described as smaller inclusions associated with polio. However, this finding has not been validated and is not currently used for diagnosis

Intracytoplasmic inclusions are characteristic of smallpox and are not seen in HSV or VZV infections. It may be difficult to distinguish smallpox, HSV, and VZV by H&E alone

‡ The presence of HPV in metastatic carcinoma to a lymph node suggests a primary in the tonsil (if in the head and neck) or cervix

AIDS-associated B-cell lymphoma, plasmablastic lymphoma, posttransplantation lymphoproliferative disorder, lymphomatoid granulomatosis, methotrexate-associated B-cell lymphoma, severe combined immunodeficiency-associated B-cell lymphoma, Wiskott-Aldrich syndrome-associated B-cell lymphoma, X-linked lymphoproliferative disorder-associated B-cell lymphoma, EBV-positive diffuse large B-cell lymphoma of the elderly, peripheral T-cell lymphoma, extranodal NK/T-cell lymphoma, nasal type, virus-associated hemophagocytic syndrome T-cell lymphoma

AI, acute inflammation

HSIL, high-grade squamous intraepithelial lesion

LMP-1 or HBcAg) or cellular proteins increased during infection

IPFDT, inflammatory pseudotumor-like follicular dendritic cell tumor

viral nucleic acids are amplified by polymerase chain reaction (PCR) and either quantified or visualized on tissue sections

LMP-1, latent membrane protein 1; LNA-1, latency-associated nuclear antigen (or LANA1)

LSIL, low-grade squamous intraepithelial lesion; PVAN, polyomavirus-associated nephropathy

SCC, squamous cell carcinoma; ss, single-stranded

Nuovo GJ: The utility of in situ-based methodologies including in situ polymerase chain reaction for the diagnosis and study of viral infections

Effect of formalin fixation and processing on immunohistochemistry

Effect of decalcification and fixation in paraffin-section immunohistochemistry

Loss of antigenicity in stored sections of breast cancer tissue microarrays

Paraffin section storage and immunohistochemistry: effects of time, temperature, fixation, and retrieval protocol with emphasis on p53 protein and MIB1 antigen

Assessment of antigen damage in immunohistochemistry, the vimentin internal control

Keratin expression in human tissues and neoplasms

Cytokeratin 7 and 20 expression in choroid plexus tumors: utility in differentiating these neoplasms from metastatic carcinomas

Coordinate expression of cytokeratins 7 and 20 defines unique subsets of carcinomas

Appraisal of the comparative utility of immunohistochemistry and electron microscopy in the diagnosis of childhood round cell tumors

Diagnostic strategies for unknown primary cancer

Bimodal frequency distribution of estrogen receptor immunohistochemical staining results in breast cancer: an analysis of 825 cases

American Society of Clinical Oncology/ College of American Pathologists Guideline Recommendations for Human Epidermal Growth Factor Receptor 2 Testing in Breast Cancer

Metastatic breast carcinoma to the stomach simulating linitis plastica

Utility of immunohistochemistry in distinguishing primary adenocarcinomas from metastatic breast carcinomas in the gastrointestinal tract

Pathology of mesothelioma

Immunohistochemical classification of amyloid in surgical pathology revisited

Imnunohistochemical characterization of a panel of 56 antibodies with normal human small intestine, colon, and breast tissues

issue is devoted to electron microscopy

Medically Important Fungi, 4 th edition: a guide to identification

Pathologic Diagnosis of Fungal Infections

Rapid diagnosis of smallpox infection and differentiation from its mimics

Application of immunohistochemistry to infections

Viruses associated with human cancer

The utility of in situ-based methodologies including in situ polymerase chain reaction for the diagnosis and study of viral infections

Smallpox: the basics