Abeloff's Clinical Oncology, 4th Edition

Part I – Science of Clinical Oncology

Section C – Diagnosing Cancer: Pathology and Laboratory Medicine

Chapter 16 – Principles of Oncologic Surgical Pathology

Peter B. Illei,William Westra


Key Method



Routine histopathologic examination is the cornerstone of cancer diagnosis.



Immunohistochemistry is an important adjunct that allows more precision in diagnosis, classification, and prognosis.



Fine-needle aspiration, when used properly, is very valuable to guide treatment decisions.

Specific Roles of Pathologic Examination of Tissues



Histologic diagnosis and classification of cancer



Assignment of pathologic grade and stage



Intraoperative consultation to direct a surgical procedure and ensure that appropriate diagnostic tissue is obtained



Triage of tissue for specialized testing, and banking for research purposes


Pathology as a medical discipline can be traced back to Renaissance Italy, when the autopsy was valued for correlating clinical history with pathologic findings.[1] During the nineteenth century, pathologists were primarily academicians who studied and taught the causes, mechanisms, and consequences of disease. Direct diagnostic application was, at best, a peripheral concern of a small group of full-time autopsy pathologists whose primary role was to confirm a diagnosis in the dead rather than to formulate a diagnosis in the living.[2] At the turn of the twentieth century, pathology became a more clinically relevant discipline for several reasons. Technical innovations in microscopy permitted detailed microscopic descriptions of tissue patterns and cell structures that, in turn, dramatically improved diagnostic capabilities. The introduction of anesthesia and aseptic techniques allowed surgeons to perform longer and more complicated procedures that required improved diagnostic skills, which, together with the invention of the freezing microtome,[3] made intraoperative consultation possible. Finally, an enlightened medical community and the public at large came to recognize that small and subtle tumors were easier to eradicate than were large and conspicuous ones.[4]

Given this new emphasis on early cancer detection and treatment, oncologic surgeons no longer could rely on their own clinical skills and gross observations to evaluate the presence and full extent of tumor growth. Instead, a new breed of physician/pathologist evolved with specialized training in the histologic characteristics of tumors. Beginning in 1926, the American College of Surgeons insisted on properly staffed hospital laboratories under the direction of physicians trained in clinical pathology and called for the mandatory systematic examination of all surgical specimens to culminate in a report detailing the pathologic findings.

The role of the modern surgical pathologist is that of a consultant clinician who is closely affiliated with many branches of medicine, including all surgical specialties, internal medicine, medical oncology, dermatology, neurology, and diagnostic and therapeutic radiology.


The surgical pathology evaluation of a presumptive tumor specimen strives to specify clearly and comprehensively the presence, nature, and extent of a tumor in a way that guides further therapy, measures results, and predicts future outcome. The report integrates the macroscopic and microscopic findings and the results of ancillary techniques such as immunohistochemistry and molecular genetic analysis. The scope and complexity of the surgical pathology report have increased dramatically over recent years, and groups such as the Association of Directors of Anatomic and Surgical Pathologists and the Cancer Committee of the College of American Pathologists recently have issued a number of standardized protocols that are organ and/or tumor specific. [5] [6] [7] [8] [9]

For any style and format, the contents of the surgical pathology report always include information regarding the presence or absence of tumor, size and location of the tumor, histologic classification, pathologic staging, adequacy of tumor removal (i.e., status of the surgical margins), and tumor grade.[10] These anatomic measurements are supplemented by relevant immunohistochemical, cytogenetic, and molecular biologic information to refine tumor classification and to guide the selection of therapeutic options.

Tumor Classification

As strategies for the treatment of oncology patients become increasingly sophisticated and individualized, current classification schemes aim to categorize tumors in ways that are precise, reproducible, and clinically relevant. The classification of tumors is hierarchical, progressing from fundamental distinctions in biologic potential (e.g., benign vs. premalignant vs. malignant) to broad separation in cellular differentiation (e.g., epithelial vs. mesenchymal vs. lymphoid vs. melanocyctic) to much finer dissection of phenotypic expression (e.g., squamous vs. glandular, chondroid vs. osteoid, B cell vs. T cell).

Standardized international nomenclature for tumor classification has been promoted successfully by a variety of professional organizations. Most notably, the Armed Forces Institute of Pathology (AFIP) continues to update its series on tumor pathology, the Atlas of Tumor Pathology, in an effort to “promote a consistent, unified, and biologically sound nomenclature; guide the surgical pathologist in the diagnosis of the various tumors and tumor-like lesions; and provide relevant histogenetic, pathogenetic, and clinicopathologic information on these entities.”[11] The World Health Organization also has implemented a program with similar aims. Its recent collection of books on the classification of tumors has taken a more deliberate approach to incorporate relevant genetic data as a component in naming and characterizing tumors.[12]

Pathologic Staging

Stage and grade traditionally have been used as parameters to characterize the clinical severity of a tumor. Stage outweighs grade in importance as an indicator of patient outcome for most cancers, with the notable exception of many soft tissue sarcomas. Stage is a measure of tumor growth. It takes into account the size of a tumor and the anatomic extent to which it has spread. Tumor staging permits valid comparison between groups of patients, facilitates exchange of information among treatment centers, guides selection of therapy, and allows empirical estimation of patient outcome.[13] The widely used TNM system uses three components to express the anatomic extent of disease: T is a measure of the local extent of tumor spread (size of tumor); N indicates the presence or absence of metastatic spread to regional lymph nodes; and M specifies the presence or absence of metastatic spread to distant sites. T, N, and M classifications are combined to provide a stage grouping. Additional histologic features also can be important in certain tumor types and may provide prognostic information (e.g., presence and amount of necrosis in osteosarcoma after neoadjuvant therapy). The TNM system is not used for lymphomas, though staging still is an important parameter to predict outcome.

Clinical staging (cTNM) defines the anatomic extent of a tumor based on clinical evidence before the initiation of treatment. It incorporates findings obtained from the physical examination, imaging studies, surgical exploration, and tissue biopsy. Clinical stage is used as a guide for the selection of primary therapy. In contrast, pathologic staging (pTNM) is based on examination of the surgically resected specimen. Assessment of pathologic stage is contingent on the recognition and removal of tumor to allow meaningful anatomic assessment of tumor origin and local extension. Thus, the pathologist's ability to render an accurate pathologic stage sometimes is compromised by the surgical approach, especially when a tumor is fragmented or removed laparoscopically. Pathologic staging is used mainly to direct adjuvant therapy, estimate prognosis, and report results, but does not invalidate the clinical stage (and vice versa).

Tumor Grading

Tumor grade is a semiquantitative measurement of histologic differentiation compared to the normal tissue from which the tumor arises. Well-differentiated tumors (grade 1) closely resemble their non-neoplastic counterparts; poorly differentiated tumors (grade 3 or 4) do not. As a rule of thumb, the more poorly differentiated a tumor, the more aggressive its behavior; however, the impact of tumor grade on tumor behavior is highly tumor-specific. For example, the expected behavior and treatment of soft tissue sarcomas is profoundly influenced by tumor grade. In contrast, most lung carcinomas are uniformly aggressive, regardless of histologic grade. For other tumor types, the utility of histologic grading falls somewhere between these two extremes.

No single uniform scheme has been established for the histologic grading of malignant neoplasms; grading schemes are tumor-specific. For some tumors, histologic classification defines tumor grade. Small cell carcinoma of the lung, anaplastic carcinoma of the thyroid, and Ewing sarcoma are, by definition, high-grade (i.e., poorly differentiated) tumors, whereas carcinoid tumor of the lung, polymorphous low-grade carcinoma of the salivary glands, and small lymphocytic lymphoma are, by definition, low-grade (i.e., well-differentiated) tumors. Some tumor types are graded based on the severity of cytologic atypia (e.g., leiomyosarcoma, renal cell carcinoma); others are graded on the basis of architectural growth patterns (e.g., adenocarcinoma of the prostate); and still others are graded by the combination of cellular atypia and architectural disarray (e.g., ductal carcinoma of the breast). Effective grading strategies strive to minimize inconsistency in application and maximize the prognostic significance of tumor stratification.


The aim of intraoperative consultation is to provide rapid assessment of the tissue and to answer specific questions raised by the surgeon. The term “intraoperative consultation” is preferred over “frozen section,” because in a significant number of cases no frozen section is performed. It is the surgical pathologist's responsibility to decide which method should be used and whether the diagnosis can be made on careful gross examination alone or on touch imprint preparations without the need for freezing the tissue, thus preserving the morphology for permanent sections. Touch preparations are made by gently touching or scraping the surface of the tissue without causing permanent damage to the rest of the tissue, whereas for frozen sections the tissue must be thinly sectioned, then frozen, and, finally, sectioned using a special microtome. To benefit fully from the diagnostic value of intraoperative consultation, the surgeon and the pathologist must communicate clearly and must have a balanced appreciation for its strengths and limitations.[14]

Requests for intraoperative consultation are appropriate in four common situations ( Box 16-1 ):[15]

Box 16-1 


Frozen section, used appropriately, is an important tool in the management of a cancer patient. Several important indications can lead to a specific action on the part of the surgeon, but inappropriate uses run the risk of compromising care.

Frozen sections should be used to establish the following:



A diagnosis or pathologic stage, and thereby determine the type and extent of the operation needed



The status of the surgical margin to determine whether a wider local excision is necessary



The presence and nature of appropriate lesional tissue to distribute fresh tissue for additional laboratory studies, such as microbiology culture, flow cytometry, or molecular assays



The adequacy of a biopsy, thereby ensuring that a definitive diagnosis will be rendered on permanent histology.

Frozen sections should not be used in the following situations:



For curiosity with no clear-cut plan to use results, as this is unnecessarily costly, may tie up pathology resources, and may result in artifacts that limit interpretation of the permanent sections.



To establish a diagnosis when there is a threat of irreparable damage to tissue architecture that would preclude a definite diagnosis on permanent histology



To document a focal microscopic finding in a large specimen, because sampling error may lead to an erroneous conclusion and, consequently, an inappropriate surgical approach.



To establish a diagnosis or determine pathologic stage to guide the type or extent of the operation. For example, the use of a frozen-section or touch preparation to detect metastatic melanoma in sentinel lymph nodes now makes possible the identification of patients who would benefit from one-step lymph node dissections.



To determine the adequacy of tumor removal. Frozen-section analysis of surgical margins provides assurance of complete tumor removal at the time of surgery and thus minimizes the need for additional operations for revisions of positive margins.



To confirm the nature of the lesion. This is important to guide fresh tissue distribution for appropriate laboratory studies, e.g., to the microbiology laboratory for culture studies, to the flow-cytometry laboratory for immunophenotypic analysis, and to the genetics laboratory for cytogenetic or molecular analysis. Confirmation can be achieved with either touch preparations or frozen sections, depending on the specimen and suspected process.



To assess the presence and quality of lesional tissue before the operation is completed. In this instance, the frozen section serves not to establish an intraoperative diagnosis but to ensure that adequate, lesional tissue has been secured such that a definite diagnosis can be rendered on permanent histologic examination.

Accurate frozen-section diagnosis is compromised by inherent limitations of the technique, [16] [17] especially suboptimal preservation of cytologic and histologic detail. Tissue fragmentation due to the presence of fat, bone, or foreign material and tissue distortion secondary to ice crystal formation may obscure the true identity of the pathologic process. Some specimens, particularly delicate tissue fragments in which important diagnostic distinctions are made on the basis of subtle architectural and cytologic alterations, are particularly susceptible to frozen section-induced morphologic alterations. In these cases, frozen-section analysis may be counterproductive and may compromise an accurate tissue diagnosis. For example, frozen-section evaluation of nonpalpable breast lesions, once a routine practice, is no longer encouraged because of a high error rate and a propensity to obscure irrevocably the subtle morphologic distinction between benign hyperplasia and intraductal carcinoma.

Accuracy of the frozen section also is limited by time constraints, so that a cursory microscopic evaluation of large specimens is highly vulnerable to sampling error. Thus, invasion within a large villous adenoma of the colorectum, transcapsular extension of an encapsulated follicular neoplasm of the thyroid, or malignant transformation within a benign mixed tumor of the parotid may elude detection when microscopic examination is limited to one or two frozen sections.

When the frozen section is used properly, its diagnostic accuracy now routinely exceeds 97% in both academic centers and general practice settings. [18] [19] Diagnostic accuracy is especially high when the frozen section is used to make broad and fundamental distinction between pathologic processes (e.g., presence of tumor vs. absence of tumor, inflammatory process vs. neoplastic process, benign tumor vs. malignant tumor). Conversely, accuracy diminishes with increasing demands for precise tumor subclassification. Regardless of the situation, accuracy is improved when the pathologist is provided with appropriate clinical information and the indication for the consult.


Immunohistochemistry (IHC) localizes specific constituents in tissues based on antibody recognition of tissue antigens. It was developed initially by Coons in 1940 as an investigative immunofluorescence technique to detect antigens in frozen tissue sections.[20] However, technical advances over the past three decades have thrust the technique into the forefront of diagnostic pathology.[21] The development of nonfluorescent chromogens has allowed visualization of antigen-antibody binding by using the conventional light microscope. The introduction of various amplification steps (e.g., peroxidase-antiperoxidase method, avidin-biotin conjugate method, polymer-based labeling system) has significantly improved sensitivity and specificity. The development of novel protocols that use two or more primary antibodies that are detected by two different detection systems has enabled us to make more accurate diagnoses, even in small biopsies with limited amount of tissue. The discovery of means to “unmask” antigens (e.g., enzyme digestion, heat-induced epitope retrieval) in formalin-fixed archival tissue has permitted consistency in staining despite variations in tissue fixation and processing. Finally, the development of techniques to manufacture highly specific monoclonal antibodies has greatly expanded the arsenal of probes that can target virtually any immunogenic marker.[22] With the adaptation of the technique to formalin-fixed and paraffin-embedded tissues, IHC now is compatible with standard tissue-processing procedures and can even be performed retrospectively on tissue blocks that have been archived for many years. IHC now is used routinely to address a range of key diagnostic questions.[23]

Is It a Tumor?

In some specific instances, markers identified by IHC can help make the fundamental distinction between a malignant tumor and some benign process. For example, the demonstration of kappa or lambda light-chain restriction by IHC is a good indicator of monoclonality in β-cell processes, and can help distinguish malignant lymphoma from an inflammatory process. In the prostate, loss of a basal cell layer, as indicated by the absence of high-molecular-weight cytokeratin and p63 immunostaining, helps to distinguish adenocarcinoma of the prostate from benign adenosis and other conditions that may mimic malignancy. [24] [25] As our understanding of the molecular genetic basis of human tumors expands, antibodies to products of oncogenes and tumor-suppressor genes hold much promise in recognizing neoplastic processes.

What Type of Tumor Is It?

The most basic application of IHC is tumor classification. For tumors that show no specific differentiation at the light-microscopic level, IHC can make critical distinctions among epithelial, mesenchymal, lymphoid, melanocyctic, and germ cell neoplasms ( Fig. 16-1 ). Even for differentiated tumors, IHC provides a detailed phenotypic description that goes well beyond the resolution capabilities of the standard light microscope. In studies emphasizing the importance of second-opinion surgical pathology, the use of IHC has been identified as a key factor resulting in major therapeutic and prognostic modifications for patients sent to large referral hospitals for oncologic surgery. [26] [27] Hematopathology is one example of a field that has become increasingly reliant on IHC, where the availability of antibodies to lineage-restricted antigens has made possible high-resolution and clinically relevant classification of hematolymphoid neoplasms.[28] IHC is having a similar impact on soft tissue sarcomas, primitive round blue cell tumors, and other areas of diagnostic pathology in which precise tumor classification is beyond the reach of conventional hematoxylin and eosin (H&E) histology. [29] [30]


Figure 16-1  Impact of immunohistochemistry on tumor classification. A, The small bowel biopsy shows sheets of histologically uniform epithelioid tumor cells in the lamina propria (H&E stain). B, Immunohistochemistry demonstrates that the tumor cells are diffusely positive for Melan-A, consistent with metastatic malignant melanoma. C, The gastric biopsy of another patient shows sheets of histologically similar-appearing tumor cells in the lamina propria (H&E stain). D, Immunohistochemistry demonstrates that the tumor cells are positive for estrogen receptors. The tumor cells are also positive for progesterone receptors, growth cystic fluid protein, and Her2/Neu, consistent with metastatic adenocarcinoma of the breast. In both examples, the differential diagnosis on H&E sections includes a primary or metastatic carcinoma and metastatic melanoma. The accurate diagnosis can be made easily if immunohistochemistry is used as an adjunct test.



What Is the Tissue of Origin for the Tumor?

Expression of some markers is so highly tissue specific that a single IHC stain can establish the most likely primary site for a neoplasm of unknown origin. Most tumors, however, display a distinctive pattern of IHC staining against a selected array of antibody probes, a pattern sometimes referred to as an “immunohistochemical profile.” For example, metastatic adenocarcinoma of the prostate may present as a tumor of unknown origin, and because most adenocarcinomas of the prostate are positive for prostate-specific membrane antigen (PSA) or prostein (p501s), prostatic origin can be readily identified by one or both of these markers in almost all cases.[31] In another application, the difficult distinction between malignant mesothelioma and peripheral lung adenocarcinoma is aided by unique IHC fingerprints when using a panel of antibody probes.[32] With the regular introduction of new antibodies into the diagnostic armamentarium, these profiles are becoming increasingly elaborate. Internet-accessible databases have become very useful in disseminating updated information regarding IHC profiles of various tumor types based on published data.[33]

Is There Evidence of Metastasis?

IHC can optimize detection of micrometastases when traditional microscopic examination is too crude to detect scattered individual tumor cells. Staining of sentinel lymph nodes with cytokeratin to detect metastatic breast carcinoma and melanocytic markers (e.g., Hmb-45, Melan-A) to detect metastatic melanoma is now used routinely for the accurate pathologic staging of regional lymph nodes. [34] [35]

What Is the Prognosis?

IHC also can help to measure determinants of disease outcome. For various tumor types, a high proliferation rate portends aggressive tumor behavior and poor outcome. The traditional practice of counting mitotic figures as a crude measure of tumor proliferation is being replaced by more quantitative assessment of proliferation activity as measured by IHC detection of certain nuclear antigens that are expressed during stages of active cell division (e.g., Ki-67).[36] Perhaps more important, IHC can help predict tumor response to certain therapies. Detection of estrogen receptors, progesterone receptors, and HER-2/neu has direct and immediate therapeutic implications for breast cancer treatment, as do prognostic indicators that now are routinely incorporated into surgical pathology reports of all invasive breast carcinomas. [37] [38] A wave of new antibody probes against oncogene products, tumor-suppressor gene products, and various cell cycle signaling proteins (e.g., activated kinases) may help individualize treatment regimens based on specific expression profiles.[39] However, the application of these markers to prognosis will require standardized technical protocols, defined cutoff values for positive results, and clinical validation studies with uniform treatment arms and adequate follow-up.[40]


In contrast to tissue histopathology with its strong reliance on architectural patterns of tumor growth, cytopathology extracts diagnostic information from the appearance of individual cells and cell clusters. Although its use has surged over the last two decades, cytopathology is hardly a new technique. Indeed, attempts to define distinctions between benign and malignant cells scraped from the surfaces of tumors constitute the origin, not the pinnacle, of contemporary diagnostic pathology.[3] George Papanicolaou (1883–1962) usually is credited with the rediscovery of cytopathologic examination. He not only demonstrated its value regarding diagnostic accuracy, but he launched its routine use as a highly effective means of reducing cancer-related morbidity and mortality.

Fine-needle aspiration (FNA) uses a fine-gauge needle to remove cells from a suggestive mass for microscopic examination. Its primary role is to guide treatment decisions, and in this role, FNA offers several significant advantages over the frozen section:



It provides a preoperative rather than intraoperative diagnosis. Some researchers estimate that up to 80% of all thyroid surgery can be avoided by routinely aspirating thyroid nodules.[41]



It is cost-effective. FNA is a simple technique that is inherently economical and often circumvents the need for a much more costly surgical intervention.



It is safe. FNA eliminates the need for general anesthesia and minimizes the risk of complications associated with more invasive procedures for tumor acquisition.

With notable exceptions (e.g., testicular masses, ovarian masses, primary malignant melanomas), FNA is no longer believed to facilitate tumor spread or induce severe hemorrhage, an occasional complication of larger-bore needles.[42]

Although reported accuracy rates for FNA range from 90% to 99%, divergent opinions persist about its reliability and its role in clinical management. Most authorities, however, would accept the following generalities:[43]



Accuracy is related to the site and nature of the neoplasm. FNA is not very useful in those situations in which tumor classification depends less on cytologic features and more on architectural patterns, such as locally invasive tumor growth. When dealing with encapsulated neoplasms of the thyroid, for example, FNA is notoriously inaccurate in separating follicular adenomas from follicular carcinomas, because a diagnosis of malignancy is contingent on the histologic demonstration of tumor invasion.



Accuracy is related to the scope of the clinical question. FNA is highly reliable when grappling with broad distinctions (e.g., presence of tumor vs. absence of tumor, inflammatory process vs. neoplastic process, benign tumor vs. malignant tumor), but accuracy diminishes for precise tumor subclassification. Moreover, largely due to the impact of limited tumor sampling by FNA, a malignant tumor is more likely to be misdiagnosed as benign (i.e., false-negative result) than a benign tumor is to be misdiagnosed as malignant (i.e., false-positive result).



Accuracy is related to the experience of the cytopathologist and the quality of the slide preparations, which, in turn, depends on the experience of the aspirator. Under ideal conditions the aspiration is performed by a cytopathologist trained to do fine-needle aspirations or, alternatively, in the presence of a cytopathologist, thus allowing an immediate assessment of the quality and quantity of the aspirate. Because of the combined impact of these factors, FNA is best used as an adjunctive diagnostic tool. It should complement, not supplant, the clinical, radiographic, and laboratory findings.


Breakneck developments in molecular biology, biotechnology, and bioinformatics are driving a molecular revolution in pathology. Contemporary research resulting in the identification of thousands of new genes and providing insight into the function and complex interaction of these genes will present the pathologist with new opportunities for solving old diagnostic problems. The technologic armamentarium of the surgical pathologist, once reliant solely on the light microscope to detect phenotypic alterations, now includes sophisticated tools to isolate and compare single cell populations in a tumor and to detect submicroscopic alterations in gene integrity, gene expression, and gene translation. [44] [45]

The direct application of molecular techniques to diagnostic oncologic pathology has just begun, but as discussed in subsequent sections, a growing number of examples bolster the bold claims that molecular analysis will profoundly aid the diagnosis, prognosis, and treatment of tumors. For hematologic neoplasms, delineation of various chromosomal translocations has allowed more clinically relevant classification of leukemias and lymphomas. [46] [47] Genetic analysis is having a similar impact on soft tissue tumors, as detection of tumor-specific translocations can be essential to classify these neoplasms.[48] Increased access to detailed genetic information also will provide a clearer picture of patient outcome and help individualize treatment plans. For hematopoietic neoplasms, detection of specific translocations already provides prognostic information independent of morphologic and immunophenotypic characterization. Moreover, the products of these transforming genes provide attractive targets for promising new therapeutic agents.

For sporadic epithelial neoplasms, the molecular genetic makeup of a tumor is not currently integrated into the inventory of more classic prognostic and predictive determinants. Efforts to do so have been stalled by the number and complexity of genetic alterations, the absence of standardized methods to measure and interpret test results, the high cost and limited availability of the technology, and the absence of well-designed clinical studies to assess clinical utility.[49] The application of such techniques is well appreciated in breast cancer, for which quantitative measures of hormone-receptor expression and HER-2/neu gene amplification now are routinely incorporated to assess outcome and guide therapy. These applications in breast cancer forecast a coming era when pharmacologic and radiation sensitivity profiles based on molecular genetic alterations will permit customized treatment of individual patients.

The molecular revolution will enhance the role of the surgical pathologist in the multidisciplinary approach to the patient with cancer, but it will not replace it. Advances in basic tumor research depend on the involvement of well-trained pathologists, not only to characterize tumors accurately with respect to site of origin and pathologic grade but also to distinguish normal and neoplastic tissue, and identify subtle degrees of morphologic changes in an individual section. With all the developments of new technology and approaches, classic light microscopy remains the cornerstone of tumor diagnostics and the starting point for the application of any new prognostic or therapeutic marker.


  1. Rosai J: Some considerations on the origin, evolution, and outlook of American surgical pathology.   In: Rosai J, ed. Guiding the Surgeon's Hand: The History of American Surgical Pathology,  Washington, DC: American Registry of Pathology, Armed Forces Institute of Pathology; 1997:1-5.
  2. Fechner RE: The birth and evolution of American surgical pathology.   In: Rosai J, ed. Guiding the Surgeon's Hand: The History of American Surgical Pathology,  Washington, DC: American Registry of Pathology, Armed Forces Institute of Pathology; 1997:7-21.
  3. Wright Jr JR: The development of the frozen section technique: the evolution of surgical biopsy, and the origins of surgical pathology.  Bull Hist Med1985; 59:295-326.
  4. Bloodgood JC: Prevention, diagnosis and treatment of cancer in its earliest stages.  South Med J1926; 19:287-292.
  5. Rosai J: Standardized reporting of surgical pathology diagnoses for the major tumor types: a proposal. The Department of Pathology, Memorial Sloan-Kettering Cancer Center.  Am J Clin Pathol1993; 100:240-255.
  6. Association of Directors of Anatomic and Surgical Pathology : Recommendations for reporting soft tissue sarcomas.  Am J Clin Pathol1999; 111:594-598.
  7. Association of Directors of Anatomic and Surgical Pathology : Recommendations for the reporting of specimens containing oral cavity and oropharynx neoplasms.  Am J Clin Pathol2000; 114:336-338.
  8. Association of Directors of Anatomic and Surgical Pathology Agency : Recommendations for the reporting of resected prostate carcinomas.  Hum Pathol1996; 27:321-323.
  9. Association of Directors of Anatomic and Surgical Pathology : Recommendations for the reporting of breast carcinoma.  Hum Pathol1996; 27:220-223.
  10. Westra WH: General approach to surgical pathology specimens.   In: Hruban RH, Westra WH, Phelps TH, Isacson C, ed. Surgical Pathology Dissection: An Illustrated Guide,  New York: Springer-Verlag; 1996:1-13.
  11. In: Rosai J, Sobin LH, ed. Atlas of Tumor Pathology, 3rd series,  Washington, DC: Armed Forces Institute of Pathology; 1991.
  12. In: Kleihues P, Sobin LH, ed. World Health Organization Classification of Tumors,  Lyon, France: IARC Press; 2000.
  13. In: Greene FL, Page DL, Fleming ID, et al ed. AJCC Cancer Staging Manual,  6th ed.. Philadelphia: Lippincott Williams & Wilkins; 2002.
  14. Ackerman LV, Ramirez GA: The indications for and limitations of frozen section diagnosis: a review of 1,269 consecutive frozen section diagnoses.  Br J Surg1959; 46:336-350.
  15. Zarbo RJ, Schmidt WA, Bachner P, et al: Indications and immediate patient outcomes of pathology intraoperative consultations: College of American Pathologists/Centers for Disease Control and Prevention Outcomes Working Group Study.  Arch Pathol Lab Med1996; 120:19-25.
  16. Westra WH, Pritchett DD, Udelsman R: Intraoperative confirmation of parathyroid tissue during parathyroid exploration: a retrospective evaluation of the frozen section.  Am J Surg Pathol1998; 22:538-544.
  17. Wick MR: Intraoperative consultations in pathology: a current perspective.  Am J Clin Pathol1995; 104:239-242.
  18. Ferreiro JA, Myers JL, Bostwick DG: Accuracy of frozen section diagnosis in surgical pathology: review of a 1-year experience with 24,880 cases at Mayo Clinic Rochester.  Mayo Clin Proc1995; 70:1137-1141.
  19. Zarbo RJ, Hoffman GG, Howanitz PJ: Interinstitutional comparison of frozen-section consultation: a College of American Pathologists Q-Probe study of 79,647 consultations in 297 North American institutions.  Arch Pathol Lab Med1991; 115:1187-1194.
  20. Coons AH, Creech HJ, Jones RN: Immunological properties of an antibody containing a fluorescent group.  Exp Biol Med1941; 47:200.
  21. Chan JK: Advances in immunohistochemistry: impact on surgical pathology practice.  Semin Diagn Pathol2000; 17:170-177.
  22. Taylor CR, Shi S-R, Barr NJ, Wu N: Techniques of immunohistochemistry: principles, pitfalls, and standardization.   In: Dabbs DJ, ed. Diagnostic Immunohistochemistry,  New York: Churchill Livingstone; 2002:3-43.
  23. Wick MR, Ritter JH, Swanson PE: The impact of diagnostic immunohistochemistry on patient outcomes.  Clin Lab Med1999; 19:797-814.vi
  24. Shah RB, Zhou M, LeBlanc , et al: Comparison of the basal cell-specific markers, 34betaE12 and p63, in the diagnosis of prostate cancer.  Am J Surg Pathol2002; 26:1161-1168.
  25. Zhou M, Aydin H, Kanane H, Epstein JI: How often does alpha-methylacyl-CoA-racemase contribute to resolving an atypical diagnosis on prostate needle biopsy beyond that provided by basal cell markers?.  Am J Surg Pathol2004; 28:39-43.
  26. Weir MM, Jan E, Colgan TJ: Interinstitutional pathology consultations. A reassessment.  Am J Clin Pathol2003; 120:5-12.
  27. Westra WH, Kronz JD, Eisele DW: The impact of second opinion surgical pathology on the practice of head and neck surgery: a decade experience at a large referral hospital.  Head Neck2002; 24:684-693.
  28. Chu PG, Chang KL, Arber DA, Weiss LM: Immunophenotyping of hematopoietic neoplasms.  Semin Diagn Pathol2000; 17:236-256.
  29. Devoe K, Weidner N: Immunohistochemistry of small round-cell tumors.  Semin Diagn Pathol2000; 17:216-224.
  30. Sheridan T, Herawi M, Illei P, Epstein J: The Role of P501S and PSA in the diagnosis of metastatic adenocarcinoma of the prostate.  Mod Pathol2006; 19:161A.
  31. Suster S: Recent advances in the application of immunohistochemical markers for the diagnosis of soft tissue tumors.  Semin Diagn Pathol2000; 17:225-235.
  32. Moran CA, Wick MR, Suster S: The role of immunohistochemistry in the diagnosis of malignant mesothelioma.  Semin Diagn Pathol2000; 17:178-183.
  33. Immunoquery [Dennis M: Frisman]: Last updated 10/27/02. Available from http://www.ipox.org
  34. Yared MA, Middleton LP, Smith TL, et al: Recommendations for sentinel lymph node processing in breast cancer.  Am J Surg Pathol2002; 26:377-382.
  35. Baisden BL, Askin FB, Lange JR, Westra WH: HMB-45 immunohistochemical staining of sentinel lymph nodes: a specific method for enhancing detection of micrometastases in patients with melanoma.  Am J Surg Pathol2000; 24:1140-1146.
  36. Brown DC, Gatter KC: Ki67 protein: the immaculate deception?.  Histopathology2002; 40:2-11.
  37. Nunes RA, Harris LN: The HER2 extracellular domain as a prognostic and predictive factor in breast cancer.  Clin Breast Cancer2002; 3:125-135.
  38. Hayes DF, Thor AD: c-ErbB-2 in breast cancer: development of a clinically useful marker.  Semin Oncol2002; 29:231-245.
  39. Griffin J: The biology of signal transduction inhibition: basic science to novel therapies.  Semin Oncol2001; 28:3-8.
  40. Seidal T, Balaton AJ, Battifora H: Interpretation and quantification of immunostains.  Am J Surg Pathol2001; 25:1204-1207.
  41. Gharib H: Fine-needle aspiration biopsy of thyroid nodules: advantages, limitations, and effect.  Mayo Clin Proc1994; 69:44-49.
  42. Amedee RG, Dhurandhar NR: Fine-needle aspiration biopsy.  Laryngoscope2001; 111:1551-1557.
  43. Ellis GL, Auclair PL: Fine-needle aspiration biopsy of salivary glands.   In: Ellis GL, Auclair PL, ed. Tumors of the Salivary Glands,  Washington, DC: Armed Forces Institute of Pathology; 1995:441.
  44. El-Naggar AK: Methods in molecular surgical pathology.  Semin Diagn Pathol2002; 19:56-71.
  45. Gabrielson E, Berg K, Anbazhagan R: Functional genomics, gene arrays, and the future of pathology.  Mod Pathol2001; 14:1294-1299.
  46. Chan JK: The new World Health Organization classification of lymphomas: the past, the present and the future.  Hematol Oncol2001; 19:129-150.
  47. Sen F, Vega F, Medeiros LJ: Molecular genetic methods in the diagnosis of hematologic neoplasms.  Semin Diagn Pathol2002; 19:72-93.
  48. Singer S: New diagnostic modalities in soft tissue sarcoma.  Semin Surg Oncol1999; 17:11-22.
  49. Jones D, Fletcher CD: How shall we apply the new biology to diagnostics in surgical pathology?.  J Pathol1999; 187:147-153.