Cancer staging is the process by which medical professionals determine the extent to which cancer has developed within the body, including the size of the primary tumor, involvement of nearby lymph nodes, and presence of distant metastasis.[1] This assessment, typically conducted after initial diagnosis, provides a standardized framework for describing the severity of the disease and is essential for predicting patient outcomes and tailoring treatment strategies.[2]The most widely used system for cancer staging is the American Joint Committee on Cancer (AJCC) TNM classification, which categorizes tumors based on three key components: T for the size and extent of the primary tumor (ranging from T0, no evidence of tumor, to T4, advanced local invasion); N for regional lymph node involvement (N0 indicates no spread, while N1–N3 denote increasing node affected); and M for distant metastasis (M0 for none, M1 for presence).[1][2] These TNM descriptors are combined into overall stage groups from 0 to IV, where stage 0 represents carcinoma in situ (abnormal cells confined to their origin without invasion); stages I and II indicate localized cancer with limited spread; stage III signifies regional extension, often to lymph nodes; and stage IV denotes advanced disease with distant metastases, generally associated with poorer prognosis.[1][2]Staging can be clinical (cTNM), derived from physical exams, imaging (such as CT or MRI scans), and biopsies before definitive treatment, or pathologic (pTNM), which incorporates surgical findings for greater accuracy, particularly after tumor resection.[2] While the AJCC TNM system applies to most solid tumors, specialized frameworks exist for certain cancers, such as the FIGO system for gynecologic malignancies, and broader summary staging (e.g., SEER's in situ, localized, regional, distant) is used for population-based research and registries.[1][2]The primary purposes of staging include guiding therapeutic decisions—such as opting for surgery in early stages or systemic therapies like chemotherapy in advanced cases—estimating survival probabilities (e.g., higher stages correlate with lower 5-year survival rates), and facilitating communication among healthcare providers and eligibility for clinical trials.[1][2] Once assigned at diagnosis, a patient's stage generally remains fixed, though restaging may occur post-treatment to evaluate response, underscoring its role in ongoing cancer management.[1]
Fundamentals of Cancer Staging
Definition and Importance
Cancer staging is the process by which medical professionals determine the extent of a cancer in a patient's body, including the size of the primary tumor and whether the cancer has spread to nearby tissues, lymph nodes, or distant sites.[1] This assessment typically classifies the cancer's progression from localized disease to advanced metastatic stages, often using standardized systems like the TNM classification to provide a structured description of its scope.[1]Staging is initially performed at the time of diagnosis through a combination of physical exams, imaging tests, laboratory analyses, and sometimes biopsies, and it may be updated after treatment to reflect changes in the disease's extent.[1] Importantly, staging focuses on the anatomical spread and size of the cancer, distinguishing it from tumor grading, which evaluates the microscopic appearance and aggressiveness of cancer cells to predict their growth behavior.[3]The primary importance of cancer staging lies in its role as a foundational tool for guiding clinical decisions and improving patient outcomes. By delineating the cancer's extent, staging informs personalized treatment strategies, such as determining whether surgery, radiation, chemotherapy, or systemic therapies are most appropriate, and helps predict prognosis through stage-specific survival rates.[1] For instance, early-stage cancers often have higher cure rates with localized interventions, while advanced stages may require multimodal approaches to manage metastasis.[4] Additionally, accurate staging ensures eligibility for clinical trials, allowing patients access to innovative therapies tailored to specific disease stages.[1]Beyond individual care, cancer staging facilitates effective communication among healthcare providers, ensuring consistent terminology and shared understanding across multidisciplinary teams.[4] It standardizes cancer classification globally, promoting uniform data collection and analysis that supports population-level surveillance programs, such as the Surveillance, Epidemiology, and End Results (SEER) program, which tracks incidence, treatment patterns, and survival trends to inform public health policies.[4] This standardization enhances the quality and comparability of cancer research worldwide, ultimately contributing to advancements in prevention, detection, and therapy.[4]
Historical Development
The development of cancer staging systems originated in the early 20th century with rudimentary, site-specific classifications aimed at predicting prognosis based on tumor extent. A seminal example is the Dukes classification for rectal cancer, introduced by British pathologist Cuthbert Dukes in 1932, which categorized disease into stages A through C according to depth of invasion and lymph node involvement, achieving notable prognostic accuracy for its time.[5] Similar early schemes emerged for other sites, such as the Whitmore-Jewett system for prostate cancer in 1975, but these varied widely in criteria and nomenclature, leading to inconsistencies that hindered comparative research and treatment planning across institutions and countries.In the mid-20th century, efforts toward standardization gained momentum with the work of French oncologist Pierre Denoix, who between 1943 and 1946 developed the foundational principles of the TNM (Tumor, Node, Metastasis) system at the Institut Gustave-Roussy, emphasizing anatomic extent as a universal framework for clinical assessment.[6] The International Union Against Cancer (UICC) formally adopted the TNM classification in 1954 as an international standard, initially applying it to select sites like breast and larynx cancers.[6] Concurrently, the International Federation of Gynecology and Obstetrics (FIGO) established its own staging system for gynecologic malignancies in 1958, starting with cervical cancer, which operated independently but was later aligned with TNM principles to facilitate global consistency.[7] In the United States, the American Joint Committee on Cancer (AJCC) was founded in 1959 to promulgate adapted staging guidelines, fostering collaboration with UICC.[8]Joint efforts between AJCC and UICC culminated in the first unified TNM edition published in 1968, marking a shift from disparate, site-specific approaches to a cohesive, evidence-based global framework that expanded to over 50 cancer types.[6] Subsequent revisions refined the system, with the 8th edition in 2017 retaining its anatomic core while incorporating prognostic refinements for specific sites. In 2021, the AJCC transitioned from fixed editions to modular "versions" to enable more frequent, site-specific updates. Version 9, with major updates released in 2024 and effective January 1, 2025, further evolves this legacy by integrating data from large-scale registries like the Surveillance, Epidemiology, and End Results (SEER) program to update criteria for multiple tumor sites, enhancing precision in staging and outcome prediction.[9] These milestones addressed the limitations of pre-TNM systems, which were fragmented and non-comparable, enabling standardized communication, clinical trials, and personalized care worldwide.
Types of Staging
Clinical Staging
Clinical staging, denoted as cTNM, involves assessing the extent of cancer prior to treatment initiation through non-invasive or minimally invasive methods to estimate tumor size, regional lymph node involvement, and distant metastasis. This process relies on physical examinations, imaging modalities such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET) scans, endoscopy, and biopsies to gather data without surgical intervention.[2] Tumor markers, like prostate-specific antigen (PSA) levels in blood for prostate cancer, may also contribute to this evaluation by providing biochemical indicators of disease extent.[2] In select cases, such as breast or melanoma, sentinel lymph node biopsy—a targeted procedure to sample the first draining lymph node—can refine nodal assessment while remaining part of pre-treatment staging.[2]The accuracy of clinical staging varies by cancer type and site, often achieving 70-90% concordance with pathologic findings in breast cancer, where overall TNM agreement reaches approximately 74%.[10] For lung cancer, inaccuracies occur in about one-third of cases compared to surgical pathology, frequently leading to understaging that underestimates disease extent.[11] These methods enable initial therapy planning, such as deciding on neoadjuvant chemotherapy or surgery, by providing a baseline prognosis without the risks of operative exploration. However, clinical staging is inherently less precise than pathologic staging, which uses surgical specimens for confirmation, with understaging reported in 20-30% of cases across various solid tumors due to limitations in imaging resolution and sampling.[2]Recent advancements in the American Joint Committee on Cancer (AJCC) 9th edition, effective in 2025, have incorporated enhanced imaging criteria to improve cTNM reliability, particularly for lung cancer through refined PET-CT interpretation of nodal FDG uptake and for head and neck cancers via updated criteria for nasopharynx and HPV-related sites.[12][13] These updates emphasize standardized imaging protocols to better align clinical assessments with prognostic outcomes.[9]
Pathologic Staging
Pathologic staging, also known as pTNM staging, determines the extent of cancer through direct examination of surgically resected tissue, providing a definitive assessment of tumor invasion, lymph node involvement, and potential distant spread. This approach relies on histopathological analysis by pathologists, who evaluate the primary tumor (pT), regional lymph nodes (pN), and distant metastasis (pM) based on microscopic findings from the surgical specimen. Unlike clinical staging, which uses non-invasive methods, pathologic staging requires surgical intervention and is typically performed after tumor resection to inform prognosis and guide adjuvant therapy decisions.[2][14]The procedures for pathologic staging begin with gross examination of the resected specimen to measure tumor size, assess margins, and identify any gross lymph node involvement or satellite lesions. Tissue samples are then processed for histologic sectioning, staining, and microscopic review to evaluate the depth of tumor invasion into surrounding tissues, the presence of cancer cells in lymph nodes, and features such as perineural or vascular invasion. Histologic grading, which assesses tumor differentiation, is integrated to provide additional prognostic context, while molecular testing—such as for mismatch repair proteins in colorectal cancer or PD-L1 expression in lung cancer—may be incorporated to identify biomarkers that influence staging and treatment. Complete surgical resection with adequate lymph node sampling (typically at least 12 nodes for many solid tumors) is crucial for reliable results, as incomplete evaluation can lead to understaging.[2][14][15]Pathologic staging offers significant advantages as the gold standard for accuracy, often upstaging 15-25% of clinical stages by detecting occult nodal metastases or deeper invasion not visible on preoperative imaging. For instance, in non-small cell lung cancer, it reduces underestimation errors from clinical methods like PET-CT, which have only 58% sensitivity for nodal involvement despite high specificity. This precision aids in tailoring postoperative therapies, such as chemotherapy for upstaged node-positive cases, and provides robust prognostic data for survival estimates. However, limitations include its inapplicability to inoperable or advanced-stage cancers where surgery is not feasible, restricting its use to approximately 20-30% of patients with resectable disease. Additionally, variability in surgical thoroughness or pathologic sampling can affect reliability, and it cannot inform neoadjuvant treatment planning since findings are only available post-resection. Pathologic staging frequently revises clinical estimates, highlighting the complementary role of both approaches.03768-6/fulltext)[14][2]In the AJCC 9th edition (2025), pathologic criteria for colon cancer have been refined to explicitly incorporate tumor deposits—discontinuous extramural foci of cancer—without requiring lymph node involvement, classifying them as N1c disease to better reflect adverse prognosis and improve stage-specific survival predictions.[16]
Restaging and Other Methods
Restaging, denoted as rTNM, is performed for recurrent cancers after a documented disease-free interval, using an "r" prefix to indicate the recurrent nature of the tumor, such as rcT1 for recurrent clinical primary tumor category.[2] This approach helps evaluate disease progression or recurrence and informs subsequent treatment strategies. In contrast, ypTNM refers to pathologicstaging following neoadjuvant therapy, such as chemotherapy administered before surgery, to assess tumor response based on surgical pathology findings.[15] For example, in esophageal cancer, ypTNM stage groups are defined distinctly from initial pathologic stages due to treatment effects, with ypStage I encompassing ypT0-2N0M0 and survival outcomes differing significantly from non-treated counterparts.[17]The 9th edition of the UICC TNM classification, effective January 2025, refines residual tumor (R) classification for lung cancer after preoperative therapy, clarifying descriptors like R0 (complete resection with adequate nodal assessment) and incorporating factors such as positive margins or pleural effusions to better reflect response to treatments including immunotherapy.[18] These updates enhance prognostic accuracy by accounting for histologic features like spread through air spaces in post-treatment pathologic evaluation.[18]Beyond traditional anatomic restaging, other methods incorporate molecular and non-invasive approaches for refined prognostic assessment. Genomic profiling, such as the Oncotype DX Breast Recurrence Score test, analyzes 21 genes in hormone receptor-positive, HER2-negative early-stage breast tumors to generate a recurrence risk score (0-100), which is integrated into the AJCC prognostic stage grouping alongside TNM to guide chemotherapy decisions.[19] Scores below 18 indicate low risk, often sparing patients unnecessary adjuvantchemotherapy, while higher scores predict greater benefit from such therapy.[19]Liquid biopsies, utilizing circulating tumor DNA (ctDNA) from blood, enable real-time monitoring of tumor dynamics without invasive procedures, aiding in staging by detecting mutations with high sensitivity in advanced stages (e.g., 100% in stage II-IV non-small cell lung cancer) and assessing minimal residual disease post-treatment.[20] In colorectal and pancreatic cancers, ctDNA levels correlate with metastasis and treatment response, with decreases signaling efficacy and elevations predicting recurrence up to months before imaging.[20] These methods support precision oncology by identifying actionable alterations for targeted therapies.Research autopsies provide a final comprehensive evaluation of disease extent post-mortem, through multiregion sampling of tumors and metastases, revealing undetected spread and refining staging understanding in cases like pancreatic or breast cancer where heterogeneity influences prognosis.[21] Such analyses uncover metastatic patterns and clonal evolution, contributing to broader insights into treatment resistance.[21]These restaging and alternative methods guide adjuvant therapy decisions by integrating response data and non-anatomic factors, advancing personalized treatment in oncology.[15]
The TNM Classification System
The American Joint Committee on Cancer (AJCC) TNM system is periodically updated, with Version 9 (formerly referred to as the 9th edition) rolling out progressively since 2024 for specific cancer sites, effective January 1 following release; staging for sites not yet updated remains based on Version 8.[9]
Primary Tumor (T) Category
The Primary Tumor (T) category in the TNM classification system describes the size, local extent, and degree of invasion of the main tumor, referred to as the primary tumor.[1] This category is determined through clinical evaluation, often using imaging techniques such as CT, MRI, or ultrasound, or via pathologic assessment following biopsy or surgical resection.[22] The T designation provides critical information on the tumor's anatomical features, helping to guide prognosis and treatment planning by indicating how far the cancer has grown locally before spreading elsewhere.[23]The T category is coded on a scale from TX to T4, with increasing numbers reflecting greater tumor size or deeper invasion into surrounding tissues. TX indicates that the primary tumor cannot be assessed, while T0 signifies no evidence of a primary tumor. Tis denotes carcinoma in situ, a pre-invasive condition where abnormal cells are confined to the epithelium without breaching the basement membrane.[24] T1 through T4 categories are primarily based on the tumor's greatest dimension—the longest single measurement of the invasive component—and its involvement of adjacent structures, though exact thresholds vary by cancer site as defined in the American Joint Committee on Cancer (AJCC) manuals. For multifocal tumors, where multiple primary lesions occur, the T category is assigned based on the largest lesion's dimensions.[25][26]Site-specific criteria refine the T category to account for anatomical differences and clinical relevance. In breast cancer, T1 encompasses tumors ≤2 cm, subdivided as T1mi (≤1 mm, often for microinvasion), T1a (>1 mm to ≤5 mm), T1b (>5 mm to ≤10 mm), and T1c (>10 mm to ≤20 mm); T2 applies to tumors >2 cm but ≤5 cm, T3 to those >5 cm, and T4 to any size with direct extension to the chest wall (T4a), skin involvement such as ulceration or satellite nodules (T4b), both (T4c), or inflammatory features (T4d).[27] For lung cancer, the 9th edition AJCC staging (effective January 2025) sets T1 for tumors ≤3 cm without invasion beyond the lobar bronchus or visceral pleura, with substages T1a (≤1 cm), T1b (>1 cm to ≤2 cm), and T1c (>2 cm to ≤3 cm); T2 includes tumors >3 cm to ≤5 cm (T2a: >3 cm to ≤4 cm; T2b: >4 cm to ≤5 cm) or those involving the main bronchus or causing atelectasis; T3 covers >5 cm to ≤7 cm or parietal pleura/chest wall invasion; and T4 denotes >7 cm or invasion of mediastinal structures, heart, or separate ipsilateral lobe nodules.[25] These variations ensure the T category captures tumor behavior relevant to each organ's anatomy.The T category directly influences local therapeutic decisions, such as selecting breast-conserving surgery (lumpectomy) for smaller T1 tumors versus more extensive mastectomy for T3 or T4 lesions with skin or chest wall involvement, thereby balancing oncologic control with quality of life.[1]
Regional Lymph Nodes (N) Category
The N category in the TNM classification system evaluates the extent of regional lymph node involvement by cancer, serving as a key indicator of local-regional disease spread and prognosis. It ranges from NX, where regional lymph nodes cannot be assessed, to N0, indicating no evidence of metastasis in regional lymph nodes, and N1 through N3, which reflect progressively greater involvement based on factors such as the number, size, location, and sometimes extranodal features of affected nodes.[1]The precise definitions of N1-N3 are site-specific to account for variations in lymphatic drainage patterns. For breast cancer, N1 denotes metastasis in 1 to 3 axillary lymph nodes or in internal mammary nodes without axillary involvement, with tumor deposits exceeding 2 mm in size; N2 involves 4 to 9 axillary nodes or clinically detected internal mammary nodes; and N3 includes 10 or more axillary nodes, infraclavicular nodes, or supraclavicular involvement.[27] In colorectal cancer, the N category focuses on the regional lymph nodebasin (pericolic or perirectal nodes), with N1 defined as metastasis in 1 to 3 nodes and N2 as involvement of 4 or more nodes, regardless of size beyond confirmed metastasis.[28]Assessment of the N category occurs through clinical (cN) methods, such as imaging (e.g., ultrasound, CT, or PET) and physical examination, or pathologic (pN) evaluation following surgical procedures like sentinel lymph node biopsy or complete lymphadenectomy. Sentinel lymph node biopsy targets the initial draining node(s) to detect early spread with minimal morbidity, particularly in breast and melanoma cases. Pathologic staging distinguishes macrometastases (>2 mm), which upstage the N category, from micrometastases (0.2 to 2 mm) and isolated tumor cells (ITCs; <0.2 mm in any dimension or ≤200 cells), with ITCs classified as pN0(i+) to indicate negligible prognostic impact.[29]Recent updates in the 9th edition of the AJCC/UICC TNM staging manual, implemented in 2025, refine the N category for head and neck cancers by incorporating grading of extranodal extension (ENE), where advanced ENE (e.g., invasion into adjacent muscle, skin, or neurovascular structures) elevates the classification to N3, enhancing prognostic precision beyond node number and size alone.[30]The N category strongly predicts recurrence risk across cancer types, with higher stages correlating to elevated rates of locoregional and distant relapse; for example, in gastric cancer, N ratios (positive nodes relative to total examined) better stratify recurrence than absolute node counts, guiding adjuvant radiation or systemic therapy decisions.[31]
Distant Metastasis (M) Category
The distant metastasis (M) category in the TNM classification system assesses the spread of cancer from the primary tumor to distant sites beyond the regional lymph nodes and tissues. M0 denotes no evidence of distant metastasis, indicating the cancer is confined to the primary site and regional nodes, while M1 indicates the presence of distant metastatic disease. Subcategories within M1 provide further granularity based on the number, location, and extent of metastases, though these vary by cancer type; for instance, in gastrointestinal cancers like colorectal carcinoma, M1a specifies metastasis confined to one distant organ or site without peritoneal involvement, M1b involves multiple distant organs or sites, and M1c indicates peritoneal metastasis regardless of other sites.[22]Detection of distant metastases relies on a combination of imaging techniques and histopathological confirmation to ensure accuracy in staging. Common imaging modalities include positron emission tomography/computed tomography (PET/CT) for detecting metabolically active lesions, computed tomography (CT) for anatomical detail, magnetic resonance imaging (MRI) for soft tissue and brain evaluation, and bone scintigraphy for skeletal involvement; biopsy of suspicious lesions is often required to confirm metastatic disease. The most frequent sites of distant metastasis across solid tumors are the liver, lungs, and bones, reflecting the hematogenous dissemination patterns of circulating tumor cells.[1][32][33]Site-specific definitions refine the M category to account for unique patterns of spread in certain cancers. In prostate cancer, M1a is defined as metastasis to non-regional lymph nodes outside the true pelvis, distinguishing it from regional nodal involvement. For melanoma, the American Joint Committee on Cancer (AJCC) 8th edition incorporates detailed M subcategories, including M1d for central nervous system (CNS) metastases, which carry a particularly poor prognosis due to limited treatment options.[34][35]The presence of any M1 subcategory generally elevates the overall cancer stage to IV, shifting treatment priorities toward systemic therapies such as chemotherapy, targeted agents, or immunotherapy to address disseminated disease. However, in oligometastatic cases—characterized by a limited number of distant sites (typically 1–5)—multimodal approaches incorporating local therapies like surgery or stereotactic radiotherapy may be pursued alongside systemic treatment to potentially improve progression-free survival, though this remains under investigation in clinical trials.[1][36]
Application to Cancer Types
Solid Tumors
The TNM classification system serves as the primary framework for staging solid tumors, which are predominantly epithelial-origin carcinomas arising in organs such as the lung, breast, and prostate. Developed collaboratively by the American Joint Committee on Cancer (AJCC) and the Union for International Cancer Control (UICC), it relies on site-specific manuals that adapt the core TNM categories to the unique anatomical and biological characteristics of each cancer type.[13] These manuals, updated in the AJCC 9th edition effective January 2025 and UICC 9th edition effective January 2026, provide over 50 distinct schemas tailored to various solid tumor sites, ensuring precise prognostic assessment and treatment planning.[37] The UICC 9th edition promotes global consistency by unifying staging criteria across anatomical regions, incorporating evidence from international databases to refine tumor extent definitions.[13]In breast cancer, the TNM system indirectly integrates biomarker status such as estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) through prognostic stage groupings that modify the anatomic stage based on these factors, allowing for more nuanced risk stratification beyond pure tumor size and nodal involvement.[38] For colorectal cancer, the N category emphasizes the number of regional lymph nodes involved, building on the legacy of the Dukes classification by incorporating quantitative nodal assessment to differentiate prognoses, with modifications in recent editions to account for tumor deposits and perforation risks.[39] Lung cancer staging in the IASLC-endorsed 9th edition TNM (2025) introduces specific criteria for subsolid nodules, including ground-glass opacities, where the solid component size determines T category to better reflect indolent lepidic growth patterns and improve early-stage detection accuracy.[12]Tumor heterogeneity in solid cancers poses significant challenges to TNM staging, as variations in genetic, histologic, and microenvironmental features within the same tumor can lead to under- or overestimation of extent, necessitating multimodal approaches like advanced imaging (e.g., PET-CT) and surgical exploration for accurate pathologic confirmation.[40] This heterogeneity often requires intraoperative staging adjustments, particularly in heterogeneous tumors like non-small cell lung cancer, where preoperative imaging may miss occult nodal involvement.[41] Staging directly influences surgical decision-making; for instance, stage I lung cancer typically warrants lobectomy over sublobar resection to achieve optimal oncologic outcomes, as evidenced by superior long-term survival rates in comparative studies.[42]
Hematologic Cancers
Hematologic cancers, including lymphomas, leukemias, and multiple myeloma, differ from solid tumors in their diffuse involvement of blood, bone marrow, and lymphoid tissues, necessitating specialized staging systems that emphasize laboratory findings, symptoms, and bone marrow assessment rather than anatomic tumor size or TNM categories. Unlike the TNM system used for solid malignancies, which focuses on tumor extent, node involvement, and metastasis, hematologic staging prioritizes factors such as peripheral blood lymphocytosis, organomegaly, anemia, and elevated lactate dehydrogenase (LDH) levels to reflect disease burden and prognosis.[43] Bone marrow involvement is a key determinant, often assessed via biopsy to quantify abnormal cell infiltration, guiding treatment intensity.[44]For lymphomas, the Ann Arbor staging system, originally developed in 1971 and now primarily used through its Lugano classification modification, categorizes Hodgkin and non-Hodgkin lymphomas into stages I through IV based on the number and location of involved lymph node regions or extranodal sites.[45] Stage I indicates involvement of a single lymph node region or lymphoid structure, stage II involves two or more regions on the same side of the diaphragm, stage III spans both sides of the diaphragm, and stage IV denotes widespread extranodal dissemination, such as in the liver or bone marrow.[46] Each stage is further subclassified with an "A" for absence or "B" for presence of systemic symptoms like unexplained fever, night sweats, or weight loss exceeding 10% of body weight.[47] The Cotswolds modification, introduced in 1990, refined this framework by incorporating designations for bulky disease (e.g., nodal mass >10 cm or mediastinal mass >1/3 thoracic diameter, denoted as "X") and clarifying extranodal site coding (e.g., "E" for limited contiguous extranodal extension).[48] The Lugano classification, adopted in 2014, further integrates modern imaging like PET/CT for more precise staging while retaining the core Ann Arbor structure.[45]In multiple myeloma, a plasma cellmalignancy, the Revised International Staging System (R-ISS), established in 2015, is the current standard for prognostic staging, categorizing patients into three stages based on serum beta-2 microglobulin and albumin levels (from the original International Staging System), plus LDH levels and high-risk cytogenetic abnormalities detected by fluorescence in situ hybridization (FISH).[49] The earlier Durie-Salmon staging system, established in 1975, assesses disease extent through three stages correlated with tumor cell mass, using radiographic evidence of bone lesions, hemoglobin levels, and serum calcium concentrations.[50] Stage I represents low tumor burden with normal hemoglobin (>10 g/dL), low M-protein levels, and fewer than three lytic bone lesions; stage II is intermediate; and stage III indicates high burden with anemia (hemoglobin <8.5 g/dL), hypercalcemia (>12 mg/dL), and multiple bone lesions.[51] This system emphasizes skeletal survey for lytic lesions and laboratory markers of renal impairment or hypercalcemia, reflecting myeloma's impact on bone marrow and end-organ function.[52]Leukemias are generally not staged in the traditional sense due to their systemic nature but are classified by subtype and risk using systems like the French-American-British (FAB) or World Health Organization (WHO) classifications, which categorize acute myeloid leukemia (AML) based on morphology, cytogenetics, and molecular features such as blast percentage in bone marrow (>20% for diagnosis).[53] For chronic lymphocytic leukemia (CLL), the Rai system (used primarily in the U.S.) divides disease into five stages: stage 0 (lymphocytosis only), stages I-II (with lymph node or spleen enlargement), and stages III-IV (with anemia or thrombocytopenia).[54] The Binet system (common in Europe) simplifies this into three stages (A-C) based on the number of involved lymphoid areas (nodes, liver, spleen) and anemia or thrombocytopenia thresholds, focusing on blood counts and physical exam findings without routine imaging.[55] These approaches highlight lab values like absolute lymphocyte count (>5,000/μL) and bone marrow infiltration to stratify risk rather than anatomic spread.[56]
Stage Grouping
Grouping Process
The grouping process in cancer staging involves combining the individual T (primary tumor), N (regional lymph nodes), and M (distant metastasis) categories from the TNM classification system into discrete overall stage groups, typically ranging from stage 0 to IV, to provide a simplified prognostic framework. This aggregation is performed using site-specific tables outlined in the AJCC Cancer Staging Manual and harmonized with UICC guidelines, where each permissible combination of TNM descriptors corresponds to a predefined stage based on empirical survival data and clinical relevance. For instance, a combination such as T1N0M0 generally groups as stage I across many solid tumors, while exceptions exist due to organ-specific biology; in breast cancer, pT3N0M0 (a tumor larger than 5 cm without nodal or distant involvement) is classified as stage IIB in the anatomic staging system. Not all possible TNM combinations are assigned to groups, as certain high-T low-N scenarios may elevate the stage to reflect poorer prognosis despite limited nodal spread.[57][23][58]Site-specific algorithms govern the grouping to account for variations in tumor behavior and treatment response, often incorporating non-anatomic factors such as tumor grade or biomarkers when they significantly impact outcomes. In prostate cancer, for example, the Gleason score (now reported via Grade Groups 1-5) and serum prostate-specific antigen (PSA) levels modify the prognostic stage; tumors with Grade Group 5 (Gleason 9-10) or PSA >20 ng/mL may group into stage III even if anatomically confined (T1-T2). Similarly, breast cancer prognostic staging integrates histologic grade (e.g., Nottingham score) alongside estrogen receptor (ER), progesterone receptor (PR), and HER2 status to refine groups beyond pure TNM, ensuring alignment with adjuvant therapy decisions. These factors are applied post-TNM assignment, using lookup tables that prioritize prognostic discrimination over strict anatomic extent.[59][58]The 9th edition of the AJCC/UICC staging system, implemented progressively from January 2025 on a site-specific basis, refines these groupings for enhanced prognostic accuracy based on large-scale international databases, such as reclassifying certain lung cancer TNM combinations (e.g., shifting T1N1 from stage IIB to IIA) to better separate survival outcomes and support precision oncology. This update, aligned with the UICC's 9th edition TNM classification published in July 2025, emphasizes evidence-based revisions, including finer distinctions between early stages like IB and IIA in non-small cell lung cancer, where prior editions showed overlapping prognoses, thereby standardizing global reporting and facilitating comparative clinical trials. Overall, the process ensures consistent, reproducible staging that aids in treatment stratification and outcome prediction across diverse cancer sites.[9][60][13]
Stage Categories and Prognosis
Cancer staging culminates in the assignment of stage groups, denoted by Roman numerals from 0 to IV, which integrate the TNM categories to provide a simplified prognostic indicator. Stage 0 represents carcinoma in situ, where abnormal cells are confined to the epithelium without invasion into underlying tissues, indicating a pre-malignant or very early condition with no risk of metastasis.[2] Stages I and II denote localized or regional disease: Stage I typically involves a small, invasive tumor confined to the primary site (e.g., T1N0M0), while Stage II encompasses larger tumors or limited lymph node involvement without distant spread (e.g., T2-3N0M0 or T1N1M0), reflecting disease that remains potentially curable with local therapies.[61] Stage III signifies advanced regional involvement, such as extensive lymph node metastasis (e.g., T any, N2-3, M0), where the cancer has spread beyond the primary site but not to distant organs, often requiring multimodal treatment.[1] Stage IV indicates metastatic disease (any T, any N, M1), with spread to distant sites like the liver or lungs, marking an incurable state focused on palliation.[1]Prognosis worsens progressively with advancing stage, as higher stages correlate with increased tumor burden and dissemination, influencing 5-year relative survival rates. For instance, in breast cancer, Stage I disease yields a 5-year survival exceeding 99%, dropping to approximately 87% for regional Stages II-III and 32% for Stage IV, based on data from the Surveillance, Epidemiology, and End Results (SEER) program.[62] These rates, derived from large population cohorts by the American Joint Committee on Cancer (AJCC) and SEER, underscore how staging stratifies risk, with early detection dramatically improving outcomes across solid tumors.[63]Stage categories directly inform treatment strategies, as outlined in National Comprehensive Cancer Network (NCCN) guidelines, tailoring interventions to disease extent for optimal efficacy. Early stages (0-I) often permit curative intent with surgery alone or combined with adjuvant therapy, such as lumpectomy for Stage I breast cancer to achieve high local control.[64] Regional Stages II-III necessitate escalated approaches like neoadjuvant chemotherapy followed by surgery and radiation to address lymph node involvement, aiming for disease eradication.[64] In contrast, Stage IV shifts to palliative systemic therapies, including targeted agents or immunotherapy, to prolong survival and manage symptoms rather than cure.[65]The AJCC's 9th edition (Version 9), effective January 2025 on a site-specific basis, refines stage groupings for enhanced risk stratification, incorporating molecular and imaging advances. These adjustments, informed by international collaborations, aim to improve the prognostic accuracy of the system without altering core TNM definitions.[9]
Special Considerations
Stage Migration
Stage migration refers to the phenomenon where improvements in diagnostic techniques lead to a redistribution of patients across cancer stages without any actual change in disease biology or treatment outcomes. This results in apparent enhancements in stage-specific survival rates, as patients with more aggressive disease are reclassified from lower to higher stages, while lower stages now include only less aggressive cases. Known as the Will Rogers phenomenon—named after the comedian's quip about relocating residents between two Oklahoma towns to improve the averageincome in both—the effect creates misleading statistics that can inflate perceived survival benefits across affected stages.[66][67]The primary causes of stage migration include advancements in imaging technologies, such as positron emission tomography (PET) and PET/CT scans, which detect metastases more sensitively than traditional methods like CT or MRI alone. For instance, the incorporation of prostate-specific membrane antigen (PSMA) PET imaging in prostate cancer staging has led to significant upstaging in up to 20-30% of cases by identifying occult nodal or distant disease. Similarly, extended surgical techniques, such as sampling more lymph nodes during procedures, contribute to upstaging by revealing previously undetected involvement; in colorectal cancer, increased nodal dissection has shifted stage distributions, with studies showing up to 15-20% of patients reclassified. Screening programs exacerbate this through lead-time bias, where early detection advances the diagnosis timeline without altering the disease course—for example, prostate-specific antigen (PSA) screening has dramatically increased the proportion of stage I diagnoses, from about 20% in the pre-PSA era to over 70% in recent decades, while reducing advanced presentations.[68][69][70][71]These shifts have notable impacts on clinical research and practice, often overestimating the efficacy of treatments in clinical trials by comparing contemporary cohorts with more accurate staging to historical ones with understaging. This complicates longitudinal comparisons of survival data, as apparent improvements may reflect diagnostic refinements rather than therapeutic advances; for example, in head and neck cancers, the adoption of PET has been linked to spurious stage-specific survival gains of 5-10% due to migration alone. In breast cancer, widespread mammography screening since the 1980s has induced a clear stage shift, with early-stage diagnoses rising from 48% to 62% of cases, contributing to inflated survival statistics in lower stages without proportional mortality reductions.[72][73]To mitigate artificial shifts, Version 9 of the American Joint Committee on Cancer (AJCC) staging manual, released in 2025, incorporates refined criteria such as updated nodal subclassifications in lung cancer and clarified definitions for tumor deposits in colorectal cancer, aiming to stabilize stage distributions and reduce migration-induced biases.[74][75]
Limitations and Future Directions
The TNM staging system, with its primary emphasis on anatomic features such as tumor size, lymph node involvement, and metastasis, often overlooks the biological heterogeneity of cancers, leading to prognostic inaccuracies. For example, certain stage I tumors can display aggressive behavior and worse outcomes than some indolent stage III tumors due to unaccounted molecular drivers.[76] This anatomic focus limits the system's ability to incorporate tumor biology, as highlighted in critiques of its failure to encompass the full spectrum of prognostic variables beyond morphology.[77]Inter-observer variability further undermines the reliability of TNM staging, particularly in imaging and pathologic assessments of tumor extent and nodal status, where agreement among clinicians can range from moderate to poor.[78] In colorectal cancer, for instance, microsatellite instability (MSI) status exemplifies how molecular markers can refine staging; MSI-high tumors in stage II disease generally confer a better prognosis than anatomic TNM alone would predict, demonstrating the need for integrated approaches.[79] Additionally, the TNM framework proves less effective for rare cancers, where insufficient data hinder robust validation and adaptation of staging criteria.[80]Looking ahead, the 8th edition and Version 9 of the AJCC, and the 8th and 9th editions of the UICC staging manuals have introduced prognostic staging that incorporates biomarkers alongside anatomic TNM to enhance accuracy and personalization.[81] The UICC 9th edition, implemented in 2025, begins this shift by promoting greater integration of non-anatomic factors, such as molecular profiles, to address longstanding limitations and support tailored therapeutic decisions.[13]Advancements in artificial intelligence for imaging analysis offer promising solutions to reduce inter-observer variability and improve staging precision, with AI algorithms demonstrating performance comparable to expert radiologists in tumor detection and segmentation across cancers like breast and lung.[82] Personalized staging models leveraging genomic data, such as those derived from MSK-IMPACT sequencing, enable integration of tumor-specific mutations to refine prognostic assessments beyond traditional TNM categories.[83] Furthermore, ongoing research into liquid biopsies facilitates dynamic staging through real-time monitoring of circulating tumor DNA, allowing for adaptive evaluation of disease evolution and treatment response.[20]