aiwiki.page
English

Carcinoma

Carcinoma is a malignant epithelial tumor whose classification, behavior, diagnosis, and treatment depend on its tissue of origin and molecular characteristics.

48 keywords6 linked from15 not yet writtenWritten by AI
CancerSarcomaLung cancerBreast cancerColorectal cance…IncidenceAge Standardizat…Social Determina…Carcinoma

Carcinoma is a malignant tumor showing differentiation toward epithelial tissue, the tissue that covers body surfaces, lines internal cavities and organs, and forms many glands. It is the largest broad category of human cancer and includes most malignancies of the breast, lung, prostate, colon, pancreas, and several other organs. Carcinomas differ from cancers of connective tissue, blood-forming tissue, and melanocytes, although distinguishing these groups sometimes requires specialized laboratory investigation.

The term identifies a tumor’s epithelial character, not a single disease or a uniform degree of severity. Some carcinomas remain localized for years, whereas others invade surrounding tissue and spread early. Their behavior depends on the organ involved, microscopic subtype, molecular alterations, and extent of disease. A related term, carcinoma in situ, describes a noninvasive epithelial neoplasm that has not crossed the boundary separating its epithelial compartment from underlying tissue.

Understanding carcinoma therefore requires several complementary forms of classification: its anatomical origin, cellular appearance, biological mechanisms, and clinical stage. These distinctions determine how a tumor is diagnosed, how its likely course is estimated, and which treatments have demonstrated benefit.

Epithelial identity and major forms

Epithelia consist of closely associated cells with specialized junctions and an organized relationship to supporting tissue. Many exhibit apical–basal polarity: one surface faces a lumen or the exterior, while the opposite surface rests on a basement membrane. Epithelial cells perform functions including protection, absorption, secretion, and transport. Malignant transformation can disrupt their normal architecture, but carcinomas often retain recognizable evidence of their original differentiation, such as glands, keratin, or characteristic structural proteins.

Adenocarcinoma shows glandular differentiation or related secretory features. It is common in the colon, prostate, pancreas, stomach, and lung. Gland formation may be obvious as rounded or irregular lumina, but poorly differentiated tumors may produce only small amounts of mucin or express molecular markers of glandular lineage. An adenocarcinoma is not necessarily derived from a pre-existing anatomical gland: some arise in surface epithelia that normally contain secretory cells or have undergone a change in differentiation.

Squamous cell carcinoma resembles stratified squamous epithelium to varying degrees. Well-differentiated examples may form keratin pearls and show intercellular bridges. It occurs in skin, the head and neck, esophagus, lung, cervix, and other sites. Basal cell carcinoma, usually arising in skin, is a distinct category that commonly invades locally but only exceptionally produces distant spread. Urothelial carcinoma arises in the specialized lining of the urinary tract, especially the bladder, and includes noninvasive papillary, flat in situ, and invasive forms.

Other categories reflect specialized differentiation. Hepatocellular carcinoma resembles liver cells; renal cell carcinomas arise from renal epithelial lineages; and neuroendocrine carcinomas combine epithelial identity with neuroendocrine features. Small-cell carcinoma of the lung is a prominent high-grade neuroendocrine example. Mixed tumors may contain more than one malignant component, as in adenosquamous carcinoma. A carcinoma with spindle-shaped cells can resemble sarcoma, requiring careful interpretation rather than classification by shape alone.

Histological labels are consequently more informative when combined with the organ of origin. “Adenocarcinoma of the colon” and “adenocarcinoma of the lung” share a broad differentiation pattern but differ in molecular biology, routes of spread, staging, and treatment. Similarly, “poorly differentiated carcinoma” describes limited recognizable specialization, not a specific primary site. The apparent differentiation of a tumor also does not establish precisely which normal cell first underwent malignant transformation; experimental lineage studies and routine diagnostic classification address related but distinct questions.

Distribution and population burden

Carcinomas account for the great majority of adult cancers, although their exact proportion depends on which tumors and reporting systems are included. They are less dominant in childhood, when leukemias, central nervous system tumors, and several embryonal malignancies make up a larger share of diagnoses. Their predominance in adults reflects, in part, the extensive surface area and lifelong renewal of epithelial tissues, their exposure to environmental agents, and the time required for successive biological alterations to accumulate.

The International Agency for Research on Cancer estimated approximately 20 million new cancer cases and 9.7 million cancer deaths worldwide in 2022. Those figures cover all cancers, not carcinoma alone. Nevertheless, the leading cancer sites illustrate carcinoma’s importance: lung cancer accounted for about 2.5 million new cases, female breast cancer about 2.3 million, and colorectal cancer about 1.9 million. Most tumors in these groups are carcinomas, although organ-based statistics also include uncommon non-epithelial malignancies.

Age strongly influences incidence. Most carcinomas become more frequent with advancing age, but age distributions vary considerably. Testicular germ-cell tumors, for example, should not be grouped with carcinoma merely because they arise in an organ. Conversely, carcinomas of the thyroid, cervix, and breast can occur well before old age. Sex-related differences reflect anatomy, hormonal influences, inherited susceptibility, and historically unequal exposures such as tobacco use.

Geographical patterns are equally heterogeneous. Cervical carcinoma remains more frequent where access to vaccination, screening, and treatment of precancerous lesions is limited. Hepatocellular carcinoma varies with chronic viral hepatitis and other liver diseases. Gastric carcinoma has marked regional variation associated partly with infection, diet, and historical living conditions. Lung carcinoma patterns closely follow tobacco exposure with a delay of decades, although it also occurs in people who have never smoked.

Population comparisons require attention to data quality and definitions. Age-standardization reduces the influence of differing population age structures, but it does not remove differences in diagnostic access or registry completeness. Basal cell carcinomas and some cutaneous squamous cell carcinomas are incompletely recorded in many registries, making a comprehensive worldwide carcinoma count difficult. Recorded incidence can also rise when diagnostic intensity increases, without an equivalent rise in life-threatening disease.

The burden extends beyond deaths. Treatment-related disability, persistent symptoms, loss of income, and unequal access to timely care contribute substantially to the consequences of carcinoma. These outcomes are influenced by social determinants of health as well as by tumor biology.

Causes and predisposing conditions

Carcinoma develops through interactions among acquired cellular alterations, inherited susceptibility, tissue environment, and external exposures. A risk factor changes the probability of disease; it does not establish the cause of an individual tumor. Many carcinomas arise without a single identifiable initiating exposure, and the same exposure can produce different molecular effects in different tissues.

Tobacco smoke is a major cause of carcinomas of the lung and several other organs, including the oral cavity, larynx, esophagus, and bladder. Its carcinogens can damage DNA, while sustained exposure alters tissue environments and promotes the expansion of abnormal cell populations. Ultraviolet radiation is central to many skin carcinomas, producing characteristic forms of DNA injury. Ionizing radiation can also cause epithelial malignancies, with risk depending on factors including dose, exposed tissue, and age at exposure. Occupational agents include asbestos, associated with lung carcinoma as well as mesothelioma, and certain aromatic amines associated with bladder carcinoma.

Infections contribute to several major forms. Persistent high-risk human papillomavirus infection causes nearly all cervical carcinomas and substantial proportions of several anogenital and oropharyngeal carcinomas. Viral E6 and E7 proteins interfere with cellular tumor-suppressive pathways. Chronic hepatitis B and hepatitis C increase hepatocellular carcinoma risk through mechanisms involving chronic injury, regeneration, and, for hepatitis B, additional direct viral effects. Helicobacter pylori infection is an established cause of noncardia gastric adenocarcinoma. Epstein–Barr virus is associated with nasopharyngeal carcinoma and a subset of gastric carcinomas.

Long-standing tissue injury can create a setting favorable to malignancy. Examples include hepatocellular carcinoma arising in cirrhosis, colorectal carcinoma associated with prolonged inflammatory bowel disease, and esophageal adenocarcinoma arising in Barrett esophagus. These associations involve repeated regeneration, altered signaling, and selection of abnormal cell populations. Chronic inflammation is not, by itself, evidence that a carcinoma will develop.

Inherited pathogenic variants explain a minority of carcinomas but can confer substantial organ-specific risks. Lynch syndrome results from inherited defects affecting DNA mismatch repair and predisposes particularly to colorectal and endometrial carcinoma. Pathogenic BRCA1 or BRCA2 variants increase risks of several epithelial cancers, including breast and ovarian carcinoma. In familial adenomatous polyposis, inherited APC disruption predisposes to numerous colorectal adenomas and a very high untreated colorectal cancer risk.

Hormonal and metabolic environments also matter. Estrogen exposure influences some breast and endometrial carcinomas, while obesity is associated with several cancer types through interacting hormonal, metabolic, and inflammatory mechanisms. Such relationships are specific to tumor type and cannot be reduced to one universal pathway.

Molecular development and clonal evolution

A carcinoma generally develops through the expansion and diversification of an abnormal cell lineage. Genetic changes, altered gene regulation, and interactions with surrounding tissue collectively allow cells to proliferate outside normal constraints. No single sequence describes every carcinoma, and a tumor may contain several genetically distinct populations by the time it becomes clinically apparent.

An oncogene promotes tumor development when abnormally activated, whereas a tumor suppressor gene normally restrains processes such as proliferation or survival. Activating changes in RAS-family genes can maintain growth signaling; alterations affecting EGFR can create a growth dependency in some lung adenocarcinomas. Loss of TP53 function impairs responses to cellular stress and DNA damage, while disruption of the RB pathway weakens control over cell-cycle progression. The consequences depend on tissue context and on accompanying alterations.

The colorectal adenoma–carcinoma sequence provides a well-studied model rather than a universal timetable. Work synthesized by Bert Vogelstein and colleagues in 1990 linked successive morphological changes with alterations involving APC-related signaling, KRAS, chromosome losses, and TP53. Subsequent research established alternative routes, including serrated pathways and mismatch-repair-deficient tumors. A colorectal carcinoma therefore need not pass through the same molecular sequence as every other colorectal carcinoma.

Genomic instability can accelerate diversification. Defective mismatch repair produces microsatellite instability, characterized by length changes in short repetitive DNA sequences. Defects in homologous recombination, including those associated with BRCA1 or BRCA2 loss, impair accurate repair of certain DNA breaks. Chromosomal instability can generate gains, losses, and rearrangements affecting many genes simultaneously. In addition, epigenetic changes can silence regulatory genes without changing their DNA sequence.

Only some alterations are functional drivers; others are passengers accumulated during cell division or exposure-related injury. Even a genuine driver is not automatically a useful treatment target. A drug must inhibit a biologically important dependency at tolerable exposure, and other pathways may compensate. The presence of a molecular alteration therefore requires clinical interpretation rather than a simple mutation-to-drug correspondence.

Evolution continues during treatment. A resistant population may already be present at low frequency or emerge through additional changes. Drug exposure can favor its expansion by suppressing sensitive cells. Tumors may also change their differentiation state or exploit signals from surrounding cells. This evolutionary behavior helps explain why a biopsy from one place and time may not capture every clinically relevant property of advanced disease, and why molecular findings can change during its course.

Local invasion and metastatic spread

The transition from an epithelial neoplasm confined to its original compartment to invasive carcinoma is a major biological boundary. Carcinoma in situ remains within the epithelial compartment without stromal invasion. It may show marked cellular atypia, but the absence of invasion distinguishes it from invasive carcinoma. Not every in situ lesion progresses, and progression rates vary substantially among organs and lesion types.

Invasive cells cross the basement membrane and enter supporting tissue. This process involves changes in adhesion, movement, interaction with extracellular matrix, and responses to neighboring cells. Proteolytic enzymes can remodel matrix barriers, but invasion is not simply a matter of dissolving tissue. Cells may migrate individually, in small groups, or as cohesive fronts, depending on tumor type and local conditions.

The tumor microenvironment includes fibroblasts, immune cells, blood vessels, extracellular matrix, and soluble signaling molecules. Fibroblasts can generate the dense fibrous reaction called desmoplasia, conspicuous in many pancreatic and breast carcinomas. New vessel formation, or angiogenesis, can support tumor growth, although some tumors also use pre-existing vessels. Immune responses may eliminate abnormal cells, restrain their expansion, or become functionally suppressed within the tumor.

Epithelial–mesenchymal transition describes a cellular program involving reduced epithelial features and increased migratory properties. Experimental studies show that partial and reversible forms of this program can contribute to carcinoma progression. It is not an obligatory, all-or-none event for every invading cell: some carcinomas spread while retaining substantial epithelial organization.

Metastasis requires further steps. Cells must enter lymphatic or blood vessels, survive transport, exit at another site, and establish growth in a foreign tissue environment. Many disseminated cells fail at these stages. Others remain dormant before producing clinically detectable disease. Common sites of distant spread include the liver, lungs, bone, and brain, but patterns differ by primary tumor. Colorectal carcinoma frequently spreads to the liver, while prostate carcinoma has a strong tendency to involve bone.

Carcinomas often spread through regional lymphatics, but this is a tendency rather than a defining rule. A positive lymph node is evidence of spread and may change stage, yet it does not imply the same prognosis across all organs. Transcoelomic spread, particularly across peritoneal surfaces, is important in tubo-ovarian and gastrointestinal carcinomas. A carcinoma found in the liver may therefore be a metastasis rather than a primary liver tumor; the distinction determines its classification and clinical interpretation.

Diagnostic pathology and identification of origin

A definitive diagnosis usually depends on examining cells or tissue obtained by biopsy, surgical removal, or cytological sampling. Imaging can identify a suspicious lesion and establish its distribution, but it generally cannot establish epithelial lineage with the same certainty as pathological examination. Sampling method matters because a small specimen may demonstrate malignancy while failing to show the full subtype, invasive boundary, or range of differentiation.

Routine histopathology assesses tissue architecture and cellular features, usually in hematoxylin-and-eosin-stained sections. Findings may include irregular infiltrating glands, nests of atypical cells, abnormal mitoses, necrosis, keratinization, or mucin production. Invasion into stroma, vessels, or nerves provides important information. Cytological specimens can diagnose many carcinomas, but isolated cells often do not preserve the architecture needed to distinguish in situ from invasive disease.

Immunohistochemistry uses antibodies to detect proteins in tissue. Cytokeratins support epithelial differentiation in many settings, while other markers help identify a lineage or likely organ of origin. TTF-1 and Napsin A can support lung adenocarcinoma; GATA3 may support breast or urothelial origin in an appropriate panel; and PAX8 is useful in evaluating several renal, thyroid, and Müllerian tumors. No individual stain is perfectly sensitive or specific, and interpretation depends on morphology, clinical information, and the pattern of multiple results.

The principal alternatives include lymphoma, melanoma, sarcoma, and certain germ-cell tumors. Some of these can resemble poorly differentiated carcinoma, while a carcinoma may lose markers normally expected in its tissue of origin. Classification therefore cannot rest on a single positive stain. When material is limited, the diagnostic process must also preserve enough tissue for clinically relevant molecular tests.

Molecular analyses can refine both classification and treatment selection. Examples include testing lung adenocarcinoma for actionable driver alterations, assessing mismatch-repair proteins or microsatellite instability in colorectal and endometrial carcinomas, and evaluating hormone receptors and HER2 in breast carcinoma. A biomarker may be diagnostic, prognostic, or predictive; these roles are different. A predictive marker indicates an association with response to a particular treatment, not a guarantee of benefit.

Sometimes metastatic carcinoma is confirmed but the primary site remains unidentified after an appropriate investigation. This is termed carcinoma of unknown primary. Immunohistochemical and molecular patterns can suggest an origin, but uncertainty may persist. Pathology reports commonly integrate histological type, grade where applicable, margins, invasion-related findings, lymph-node results, and selected biomarkers rather than relying on the word “carcinoma” alone.

Stage, grade, and prognostic interpretation

Cancer staging describes the extent of disease, whereas grade describes selected microscopic features associated with biological behavior. These dimensions overlap imperfectly. A well-differentiated carcinoma can be advanced, and a high-grade carcinoma can be small and localized. Neither term should be used as a substitute for the other.

The TNM staging system is widely used for solid malignancies. T describes features of the primary tumor, N regional lymph-node involvement, and M distant metastasis. The meaning of a particular T or N category depends on the organ. Size is central in some settings, depth of invasion in others, and involvement of specified adjacent structures in still others. Stage groups commonly run from I to IV, with stage 0 used for certain in situ lesions, but exact groupings are disease-specific.

Clinical staging uses information available before definitive treatment, including examination, imaging, and biopsy. Pathological staging incorporates findings from surgical specimens. Separate conventions describe staging after preoperative treatment because therapy can alter the tumor and lymph nodes. These categories are not interchangeable: a small residual lesion after systemic therapy does not necessarily indicate that the original cancer had been biologically limited.

Grading systems also differ. Prostate carcinoma uses Gleason patterns and Grade Groups based on glandular architecture. Breast carcinoma grading evaluates tubule formation, nuclear features, and mitotic activity. Some other carcinomas use distinct two-tier or three-tier systems. “High grade” consequently has meaning within a specified classification rather than as a uniform measurement across all cancers.

Additional prognostic factors include lymphovascular invasion, perineural invasion, margin status, molecular subtype, and response to treatment. In breast carcinoma, for example, receptor status and selected genomic assays can refine risk estimates in defined populations. In oropharyngeal squamous cell carcinoma, HPV-associated disease has a separate staging framework from HPV-unrelated disease. These distinctions show why the anatomical extent alone is sometimes insufficient.

Survival statistics require careful interpretation. Overall survival measures death from any cause, while cancer-specific survival and relative survival address different questions. Estimates are based on groups treated during specified periods and do not determine an individual outcome. Earlier diagnosis can lengthen measured survival from diagnosis through lead-time effects even without postponing death. Comparisons also depend on age, coexisting illness, treatment availability, and the completeness of follow-up. Prognosis is therefore a conditional estimate grounded in a particular disease definition and population, not a fixed property of the label carcinoma.

Treatment principles and variation among sites

There is no single treatment for carcinoma. Management depends on primary site, subtype, stage, molecular findings, previous treatment, and the person’s general condition. Local treatments act mainly on a defined anatomical region, while systemic treatments reach tumor cells through the circulation. Their roles may be curative, disease-controlling, or directed toward relief of symptoms.

Surgery is a principal treatment for many localized carcinomas. The operation may remove the primary tumor, surrounding tissue, and selected regional lymph nodes. Pathological examination of the specimen establishes whether tumor reaches the cut edges and provides additional staging information. The extent of surgery differs widely: a superficial lesion may be amenable to local or endoscopic removal, whereas an invasive carcinoma may require removal of part or all of an organ.

Radiation therapy damages cellular DNA and can be used alone or with surgery and systemic treatment. It is central to treatment of many head-and-neck, cervical, prostate, and other carcinomas. Concurrent chemotherapy can increase tumor sensitivity to radiation in selected settings. Modern planning seeks to deliver an effective tumor dose while limiting exposure to adjacent organs, but adverse effects remain dependent on the treated site and dose.

Chemotherapy includes drugs that interfere with processes such as DNA replication and cell division. It may be administered before local treatment, termed neoadjuvant therapy, or afterward as adjuvant therapy to reduce recurrence risk. In advanced disease it can reduce tumor burden and prolong survival in susceptible tumor types. Endocrine treatments act through hormone-related pathways and are especially important in hormone-receptor-positive breast carcinoma and prostate carcinoma.

Targeted cancer therapy exploits defined molecular vulnerabilities. Established examples include HER2-directed treatments in selected breast and gastric carcinomas, EGFR inhibitors in appropriately selected lung carcinomas, and PARP inhibitors in specified settings involving defective DNA repair. Antibody–drug conjugates combine recognition of a cell-surface target with delivery of a cytotoxic payload. These approaches require attention to target expression, resistance mechanisms, and treatment-specific toxicities.

Cancer immunotherapy can restore or amplify antitumor immune activity. An immune checkpoint inhibitor blocks inhibitory pathways such as PD-1, PD-L1, or CTLA-4 and has established roles in multiple carcinomas. Responses are not universal; their likelihood varies by disease and biomarkers such as mismatch-repair deficiency or PD-L1 expression in defined contexts. Immune-mediated adverse effects can involve several organs.

Treatment outcomes include more than tumor shrinkage. Symptoms, physical function, and treatment burden are also measured in clinical studies. Palliative care addresses symptoms and other serious-illness needs and can accompany disease-directed treatment, rather than being limited to the final phase of life.

Prevention, precursors, and early detection

Carcinoma prevention operates at several levels: reducing causal exposures, preventing oncogenic infections, identifying selected precursor lesions, and detecting established cancers before they become symptomatic. These approaches differ in mechanism and in the strength of evidence for particular organs. There is no universal test or prevention strategy applicable to every carcinoma.

Tobacco control has a major role because smoking causes several common carcinomas. Vaccination against high-risk HPV types prevents infections responsible for cervical and other HPV-associated malignancies, while hepatitis B vaccination reduces the burden of chronic infection and associated hepatocellular carcinoma. Control of occupational carcinogens and ultraviolet exposure addresses other established causes. Such interventions act on risks over populations and often produce their full effects only after substantial delays.

A precursor lesion is not simply a small invasive cancer. Colorectal adenomas, some serrated lesions, cervical intraepithelial neoplasia, and certain other epithelial abnormalities can precede invasive carcinoma, but their probabilities of progression differ. Removal or treatment of selected precursors can interrupt a causal pathway. By contrast, a marker of elevated risk may not itself be the lesion from which cancer develops. This distinction is important when interpreting findings in breast and other tissues.

Cancer screening examines people without relevant symptoms using a defined program. Cervical screening can identify high-risk HPV infection or epithelial abnormalities, depending on the method. Colorectal screening may detect bleeding, neoplastic lesions, or both; colonoscopy also permits removal of many precursor polyps. Mammography detects some breast carcinomas before clinical presentation, and low-dose computed tomography reduces lung cancer mortality in defined high-risk populations studied in trials. Each program has its own eligibility criteria, intervals, and follow-up requirements.

Screening performance depends on test accuracy, disease frequency, participation, and access to diagnostic confirmation and effective treatment. A positive screening test is not itself a carcinoma diagnosis. False-positive findings can lead to additional procedures, while false-negative results can delay recognition. The balance of benefit and harm is therefore assessed for a complete screening pathway, not only for the initial test.

Overdiagnosis means detecting a genuine cancer that would not have caused symptoms or death during the person’s lifetime. It differs from a false-positive result, because the pathological diagnosis may be correct. Slow-growing lesions and competing causes of death contribute to this phenomenon. Increasing the number of detected carcinomas does not automatically demonstrate a reduction in mortality.

Blood-based multicancer detection methods remain an active research area. Their clinical value depends on evidence about downstream investigations, missed cancers, overdiagnosis, and mortality, rather than detection accuracy alone.

Historical development and research approaches

The terminology of cancer long predates the cellular understanding of carcinoma. Ancient Greek medical writings used terms related to a crab to describe some tumors, and Latin medical usage preserved related terminology. Those descriptions did not correspond precisely to present-day microscopic categories. The modern distinction between epithelial malignancies and other tumors required the development of microscopy and cellular pathology during the nineteenth century.

Rudolf Virchow’s work helped establish disease as a process rooted in cells, while later histological investigation clarified the organization and origins of tumors. Recognition of epithelial growth patterns, stromal invasion, and metastatic similarity gradually made carcinoma a reproducible pathological category. During the twentieth century, improved fixation, staining, surgical sampling, and organ-specific classifications allowed increasingly detailed separation of histological subtypes.

Research on environmental carcinogenesis supplied evidence that particular exposures could cause epithelial malignancies. In 1775, Percivall Pott described the association between chimney sweeping and scrotal cancer, an early occupational cancer observation. In 1915, Katsusaburō Yamagiwa and Kōichi Ichikawa reported experimental production of skin cancer in rabbits by repeated coal-tar application. These findings helped connect prolonged chemical exposure with tumor development, although the molecular mechanisms were not then understood.

The discovery of oncogenes, tumor suppressor genes, and DNA-repair pathways transformed carcinoma research in the late twentieth century. Bert Vogelstein’s colorectal studies connected molecular changes with recognizable pathological progression. Research establishing the causal role of HPV in cervical cancer created a route from basic virology to preventive vaccines. Subsequent genome-scale studies demonstrated both recurrent pathways shared across organs and major differences among tumors with superficially similar appearances.

Current experimental approaches include genetically engineered animals, patient-derived tumor grafts, and organoids, three-dimensional cultures that reproduce selected features of epithelial tissues. Each captures different aspects of disease. Organoids can preserve some tumor-specific characteristics but may lack a complete immune and vascular environment. Animal models permit investigation of invasion and treatment response but do not reproduce every feature of human disease.

Single-cell and spatial methods examine how malignant cells and surrounding tissue vary within a tumor. Analyses of circulating tumor DNA can identify tumor-derived molecular changes in blood and are established for selected molecular-testing applications. Their use for detecting minimal residual disease or directing treatment is being evaluated in disease-specific studies. Major research questions include which precursor lesions will progress, how dormant disseminated cells resume growth, and how molecular measurements can improve outcomes rather than merely identify additional abnormalities.