key: cord-021494-9glqvzfx
authors: Funkhouser, William K.
title: Pathology: The Clinical Description of Human Disease
date: 2012-07-27
journal: Essential Concepts in Molecular Pathology
DOI: 10.1016/b978-0-12-374418-0.00011-6
sha: 
doc_id: 21494
cord_uid: 9glqvzfx

nan

radiographic examination, fluid tests (blood, urine, sputum, stool), and possibly tissue biopsy.

Radiographically, abnormalities in abundance, density or chemical microenvironment of tissues allows distinction from surrounding normal tissues. Traditionally, the absorption of electromagnetic waves by tissues led to summation differences in exposure of silver salt photographic film. Tomographic approaches such as CT (1972) and NMR (1973) complemented summation radiology, allowing finely detailed visualization of internal anatomy in any plane of section. In the same era, ultrasound allowed visualization of tissue with density differences, such as a developing fetus or gallbladder stones. More recently, physiology of neoplasms can be screened with positron emission tomography (PET, 1977) for decay of short half-life isotopes such as fluorodeoxyglucose. Neoplasms with high metabolism can be distinguished physiologically from adjacent low-metabolism tissues, and can be localized with respect to normal tissues by pairing PET with standard CT. The result is an astonishingly useful means of identifying and localizing new space-occupying masses, assigning a risk for malignant behavior and, if malignant, screening for metastases in distant sites. This technique is revolutionizing the preoperative decisionmaking of clinical teams, and improves the likelihood that patients undergo resections of new mass lesions only when at risk for morbidity from malignant behavior or interference with normal function.

Pathologically, disease is diagnosed by determining whether the morphologic features match the set of diagnostic criteria previously described for each disease. Multivolume texts are devoted to the gross and microscopic diagnostic criteria used for diagnosis, prognosis, and prediction of response to therapy. Pathologists diagnose disease by generating a differential diagnosis, then finding the best fit for the clinical presentation, the radiographic appearance, and the pathologic (both clinical lab and morphologic) findings. Logically, the Venn diagram of the clinical, radiologic, and pathologic differential diagnoses should overlap. Unexpected features expand the differential diagnosis and may raise the possibility of previously undescribed diseases. For example, Legionnaire's disease, human immunodeficiency virus (HIV), Hantavirus pneumonia, and severe acute respiratory syndrome (SARS) are examples of newly described diagnoses during the last 30 years. The mental construct of etiology (cause), pathogenesis (progression), natural history (clinical outcome), and response to therapy is the standard approach for pathologists thinking about a disease. A disease may have one or more etiologies (initial causes, including toxins, mutagens, drugs, allergens, trauma, or genetic mutations). A disease is expected to follow a particular series of events in its development (pathogenesis), and to follow a particular clinical course (natural history). Disease can result in a temporary or lasting change in normal function, including patient death. Multiple diseases of different etiologies can affect a single organ, for example, infectious and neoplastic diseases involving the lung. Different diseases can derive from a single etiology, for example, emphysema, chronic bronchitis, and small cell lung carcinoma in long-term smokers. The same disease (for instance, emphysema of the lung) can derive from different etiologies (emphysema from a-1-antitrypsin deficiency or cigarette smoke).

Modern diagnostic pathology practice hinges on morphologic diagnosis, supplemented by histochemical stains, immunohistochemical stains, cytogenetics, and clinical laboratory findings, as well as the clinical and radiographic findings. Sections that meet all of these criteria are diagnostic for the disease. If some, but not all, of the criteria are present to make a definitive diagnosis, the pathologist must either equivocate or make an alternate diagnosis. Thus, a firm grasp of the diagnostic criteria and the instincts to rapidly create and sort through the differential diagnosis must be possessed by the diagnostic pathologist.

The pathologic diagnosis has to make sense, not only from the morphologic perspective, but from the clinical and radiographic vantage points as well. It is both legally risky and professionally erosive to make a clinically and pathologically impossible diagnosis. In the recent past, limited computer networking meant numerous phone calls to gather the relevant clinical and radiographic information to make an informed morphologic diagnosis. For example, certain diseases such as squamous and small cell carcinomas of the lung are extremely rare in nonsmokers. Thus, a small cell carcinoma in the lung of a nonsmoker merits screening for a nonpulmonary primary site. Fortunately for pathologists, computing and networking technologies now allow us access to preoperative clinical workups, radiographs/reports, clinical laboratory data, and prior pathology reports. All of these data protect pathologists by providing them with the relevant clinical and radiographic information, and protect patients by improving diagnostic accuracy. Just as research scientists ". . . ignore the literature at their peril. . .", diagnostic pathologists ". . . ignore the presentation, past history, workup, prior biopsies, and radiographs at their peril. . . ."

There are limitations to morphologic diagnosis by H&E stains. First, lineage of certain classes of neoplasms (including small round blue cell tumors, clear cell neoplasms, spindle cell neoplasms, and undifferentiated malignant neoplasms) is usually clarified by immunohistochemistry, frequently by cytogenetics (when performed), and sometimes by electron microscopy. Second, there are limitations inherent in a snapshot biopsy or resection. Thus, the etiology and pathogenesis can be obscure or indeterminate, and rates of growth, invasion, or timing of metastasis cannot be inferred. Third, the morphologic changes may not be specific for the underlying molecular abnormalities, particularly the rate-limiting (therapeutic target) step in the pathogenesis of a neoplasm. For example, Ret gain of function mutations in a medullary thyroid carcinoma will require DNA level screening to determine germline involvement, familial risk, and presence or absence of a therapeutic target. Fourth, the same morphologic appearance may be identical for two different diseases, each of which would be treated differently. For example, there is no morphologic evidence by H&E stain alone to distinguish host lymphoid response to hepatitis C viral (HCV) antigens from host lymphoid response to allo-HLA antigens in a liver allograft. This is obviously a major diagnostic challenge when the transplant was done for HCV-related cirrhosis, and the probability of recurrent HCV infection in the liver allograft is high.

Paraffin section immunohistochemistry has proven invaluable in neoplasm diagnosis for clarifying lineage, improving diagnostic accuracy, and guiding customized therapy. If neoplasms are poorly differentiated or undifferentiated, the lineage of the neoplasm may not be clear. For example, sheets of undifferentiated malignant neoplasm with prominent nucleoli could represent carcinoma, lymphoma, or melanoma. To clarify lineage, a panel of immunostains is performed for proteins that are expressed in some of the neoplasms, but not in others. Relative probabilities are then used to lend support (rule in) or exclude (rule out) particular diagnoses in the differential diagnosis of these several morphologically similar undifferentiated neoplasms. The second role is to make critical distinctions in diagnosis that cannot be accurately made by H&E alone. Examples of this would include demonstration of myoepithelial cell loss in invasive breast carcinoma but not in its mimic, sclerosing adenosis ( Figure 11 .1), or loss of basal cells in invasive prostate carcinoma (Figure 11 .2). The third role of immunohistochemistry is to identify particular proteins, such as nuclear estrogen receptor (ER) (Figure 11 .3) or the plasma membrane HER2 proteins (Figure 11.4) , both of which can be targeted with inhibitors rather than generalized systemic chemotherapy. Morphology remains the gold standard in this diagnostic process, such that immunohistochemical data support or fail to support the H&E findings, not vice versa.

Probability and statistics are regular considerations in immunohistochemical interpretation, since very few antigens are tissue-specific or lineage-specific. Cytokeratin is positive in carcinomas, but also in synovial and epithelioid sarcomas. This example may imply aspects of the lineage of these two sarcomas that may be helpful in our categorization of these neoplasms. Another example would be the diagnosis of small cell carcinoma in the lung of a nonsmoker. Because lung primary small cell carcinoma is extremely uncommon, in non-smokers, this diagnosis would prompt the pathologist to inquire about screening results for other, nonpulmonary, primary sites. Likewise, immunohistochemistry results are always put into the context of the morphologic, clinical, and radiographic findings. For example, an undifferentiated CD30(þ) neoplasm of the testis supports embryonal carcinoma primary in the testis, whereas a lymph node effaced by sclerotic bands with admixed CD30(þ) Reed-Sternberg cells supports nodular sclerosing Hodgkin's disease. A wealth of information is conveyed to a service pathologist in a tried-and-true H&E section. Analogous to the fact that a plain chest X-ray is the sum total of all densities in the beam path, the morphologic changes in diseased cells and tissues are the morphologic sum total of all of the disequilibria in the abnormal cells. For most neoplastic diseases, morphologic criteria are sufficient to predict the risk of invasion and metastasis (the malignant potential), the pattern of metastases, and the likely clinical outcomes. For example, the etiology and pathogenesis in small cell lung carcinoma can be inferred (cigarette smoking, with carcinogeninduced genetic mutations) and the outcome predicted (early metastasis to regional nodes and distant organs, with high probability of death within 5 years of diagnosis). New molecular data for both neoplastic and non-neoplastic diseases will most likely benefit unaffected individuals by estimating disease risk, and will most likely benefit patients by defining the molecular subset for morphologically defined diagnostic entities, thus guiding individualized therapy.

Diagnostic pathology will continue to use morphology and complementary data from protein (immunohistochemical) and nucleic acid (cytogenetics, in situ hybridization, DNA sequence, and RNA abundance) screening assays. New data will be integrated into the diagnostic process by reducing the cost and turnaround time of current technologies, and by development of new technologies, some of which are described.

For transplant candidates, major histocompatibility complex (MHC, HLA in human) screening is evolving from cellular assays and serology toward sequencing of the alleles of the class I and II HLA loci. Rapid sequencing of these alleles in newborn cord blood would allow databasing of the population's haplotypes, facilitating perfect matches for required bone marrow or solid organ transplants.

Current uses of in situ hybridization to screen for viruses (such as EBV), light chain restriction (in B lymphomas), and copy number variation (for instance, HER2 gene amplification) demonstrate the benefit of in situ nucleic acid hybridization assays. It is possible that interphase FISH/CISH will become rapid enough to be used in the initial diagnostic workup of certain patients, including for sarcomaspecific translocations, ploidy analysis in hydatidiform moles, and gene amplification of receptor tyrosine kinase genes.

Current uses of nucleic acid screening for bcr-abl translocation, donor:recipient ratios after bone marrow transplant, microsatellite instability, quantitative viral load (for EBV, BK, CMV, and others), and single gene mutations (for CFTR, Factor 2, a-1-antitrypsin) demonstrate the benefit of nucleic acid screening in diagnosis and management. It is possible that each new neoplasm will be promptly defined as to ploidy, translocations, gene copy number differences, DNA mutations, and RNA expression cluster subset, allowing residual disease screening as well as individualized therapy.

Current uses of morphology, immunohistochemistry, and molecular pathology demonstrate their benefit through improved diagnostic accuracy. However, diagnosis, extent of disease, and molecular subsets are currently imperfect estimators of prognosis and response to therapy. Relational databases which correlate an individual's demographic data, family history, concurrent diseases, morphologic features, immunophenotype, and molecular subset, and which integrate disease prevalence by age, sex, and ethnicity using Bayesian probabilities, should improve accuracy of prognosis and prediction of response to therapy. As risk correlates are developed, it is possible that healthy individuals will be screened and given risk estimates for development of different diseases.

Current uses of normal ranges for serum chemistry assumes a similar bell-curve distribution across ages, sexes, and races. This may be true for most but not all analytes. Computer reference databases will likely generate normal ranges specific for the particular age/sex/ethnicity of individual patients. Similarly, familial risk for an inherited disease may vary by ethnic group, and this variation should be used in Bayesian calculations to define risk for unaffected at-risk family members.

Current uses of liver and renal impairment to guide drug dosage demonstrate the benefit of using patient physiology to customize therapy. It is likely that individual differences in enzymatic metabolism of particular drugs (for instance, warfarin or tamoxifen) will be defined at the enzyme sequence level, and that gene haplotype data will be determined for new patients prior to receipt of these drugs.

Current uses of prostate-specific antigen (PSA) to screen for prostate carcinoma and its recurrence demonstrates the benefit of serum biomarkers in common neoplasms. It is likely that high-sensitivity screening of single and clustered serum analytes will lead to improved methods for early detection and persistence of neoplasms, autoimmune diseases, and infections.

Pathologists consider each disease to have a natural, mechanical, physicochemical basis. Each disease has an etiology (initial cause), a pathogenesis (stepwise progression), and a natural history with effects on normal function (clinical outcome). Pathologists collect the data needed to answer patients' and clinicians' questions, simply phrased as "what is it?" (diagnosis), "how it going to behave?" (prognosis), and "how do I treat it?" (prediction of response to therapy). Instincts and diagnostic criteria, as well as the optical, mechanical, chemical, and computing technologies described previously, are the basis for modern service pathology. As the human genome is deciphered, and as the complex interactions of cellular biochemistry are refined, risk of disease in unaffected individuals will be calculable, disease diagnosis will be increasingly accurate and prognostic, and molecular subsets of morphologically defined disease entities will be used to guide customized therapy for individual patients. It is a great time in history to be a pathologist.

n Clinically, diseases present to front-line physicians as patients with sets of signs and symptoms. Symptoms are the patient's complaints of perceived abnormalities. Signs are detected by examination of the patient. The clinical team (including the pathologist) evaluate the patient based on the possible causes of the signs and symptoms (the differential diagnosis). n Pathologically, disease is diagnosed by determining whether the morphologic features match the set of diagnostic criteria previously described for each disease. Pathologists diagnose disease by generating a differential diagnosis, then finding the best fit for the clinical presentation, the radiographic appearance, and the pathologic (both clinical lab and morphologic) findings. n Etiology describes the causes of a disease. One disease entity can have more than one etiology, and a single etiology can lead to more than one disease. For example, emphysema, chronic bronchitis, and small cell lung carcinoma can all occur in longterm smokers (different diseases derived from a single etiology). Likewise, the same disease (for instance, emphysema of the lung) can derive from different etiologies (emphysema from a-1-antitrypsin deficiency or cigarette smoke). n The pathogenesis of a disease describes its stepwise progression after initiation in response to a specific etiologic factor (or factors). Pathogenesis can refer to the changes in the structure or function of an organism at the gross/clinical level, and it can refer to the stepwise molecular abnormalities leading to changes in cellular and tissue function. n The natural history of a disease describes the expected course of disease, including chronicity, functional impairment, and survival. However, not all patients with a given disease will naturally follow the same disease course, so differences in patient outcome do not necessarily correspond to incorrect diagnosis. Variables that independently correlate with clinical outcome differences are called independent prognostic variables, and are assessed routinely in an effort to predict the natural history of the disease in the patient. n Likewise, variables that independently correlate with response to therapy are called independent predictive variables, and are assessed routinely in an effort to optimize therapeutic response for each patient.

Localization of antigen in tissue cells; improvements in a method for the detection of antigen by means of fluorescent antibody

General nature of the genetic code for proteins

An immunoglobulin heavy chain variable region gene is generated from three segments of DNA: VH, D and JH

In search of the origins of modern surgical pathology

Formation and detection of RNA-DNA hybrid molecules in cytological preparations

Expression in Escherichia coli of chemically synthesized genes for human insulin

Contagiousness of puerperal fever

Berliner Klinische Wochenschrift

Continuous cultures of fused cells secreting antibody of predefined specificity

Initial sequencing and analysis of the human genome

The immunology of transplantation

Specific enzymatic amplification of DNA in vitro: the polymerase chain reaction

The RNA code and protein synthesis

Ignac Semmelweis and the etiology of fetal and neonatal sepsis

Product differentiation by analysis of DNA melting curves during the polymerase chain reaction

The continuing role of morphology in the molecular age

DNA sequencing with chainterminating inhibitors

Fluorescence detection in automated DNA sequence analysis

Rudolf Virchow-father of cellular pathology

The sequence of the human genome

Molecular structure of nucleic acids; a structure for deoxyribose nucleic acid