These patients may show enlarged abdominal mass or abdominal swelling due to tumor metastasis to the liver or malignant peritoneal effusion. Vaginal bleeding due to metastasis to the endometrium has been reported in premenopausal women with AGCs [ ]. Median survival for metastatic or unresectable disease is approximately 8 to 10 months [ ]. An upper gastrointestinal tract radiography and endoscopy with biopsy have been used as gold standard tests for the detection of gastric cancer.
However, in some cases, the differentiation between a benign ulcer from a malignant one or gastric lymphoma can be challenging for radiologists in regard to subtle radiologic findings. Therefore, endoscopy with histologic confirmation has been a choice of procedure for evaluation of gastric cancer. The diagnostic accuracy of the biopsies usually increases with the increased numbers of sample taken. Many endoscopists generally take eight to ten biopsies and a minimum of six biopsies from any lesions is highly recommended with one from each quadrant and two from the center of the lesion [ ].
Biopsy should be taken from the edge of an ulcerative lesion not from the base because when the biopsy is taken from the base of the ulcer, only necrotic tissue may be obtained. Gastric forcep biopsy may have limitation for the proper diagnosis and determination of degree of differentiation in some cases. Takao et al investigated the discrepancy rates of diagnoses between biopsy samples and resection specimens and found 1. Therefore, they suggested endoscopic features should be considered together with the biopsy diagnosis to determine an appropriate treatment strategy for the lesions.
Although a great effort has been made in search of specific markers that would enable for early detection of gastric carcinoma including CEA, CA Once a diagnosis of gastric carcinoma has been made, endoscopic ultrasonography and computed tomography CT scan are usually employed for tumor staging. EUS is particularly useful to estimate the depth of tumor invasion for local staging. However, differentiating T2 and T3 gastric carcinoma may be difficult in some cases due to associated fibrosis in T2 mimicking T3 lesions.
CT scan has been used for identifying distant metastasis to lung, liver, bone, etc.
ESMO Consensus Guidelines for the Management of Patients with Metastatic Colorectal Cancer | ESMO
However, currently, MRI is limitedly used when the patients have an allergy to iodine contrast or renal failure due to motion induced artifacts, longer scanning times and a higher cost than CT scan. AGCs may show various gross appearances with different growth patterns. For standardization of the common morphologic features of AGCs, several classification systems have been proposed. The most widely used classification for macroscopic appearance of AGC is Bormann classification, dividing AGCs into four types [ ]:.
Type 1 Polypoid: Well circumscribed polypoid tumors Fig. Type 2 Fungating: Fungating tumors with marked central infiltration Fig. Type 3 Ulcerated: Ulcerated tumors with infiltrative margins Fig. Type 4 Infiltrating: Diffusely infiltrated tumors Fig. The most common macroscopic type is a type 2 fungating tumor, which are frequently located in the lesser curvature of antrum.
In contrast, polypoid type 1 and ulcerated type 3 types are commonly found in the greater curvature of corpus. On cut surface, AGC presents as a gray-white to yellow-white solid mass with a firm to hard consistency and contains areas of hemorrhage and necrosis.
Bormann type 1. Polypoid carcinoma of the stomach is located in the antrum of the lesser curvature. This elevating solid mass shows focal superficial hemorrhage. Bormann type 2. Fungating carcinoma of the stomach with extensive central ulceration involves the antrum. Bormann type 3. Ulcerated carcinoma of the stomach with infiltrative and heaped-up margins is present in the lower body of the lesser curvature.
Bormann type 4. Linitis plastica, diffusely infiltrating carcinoma of the stomach with thickening of gastric rugae involves the whole stomach. Because of heterogeneity and complexity in the morphologic characteristics of gastric carcinoma, many histologic classification systems have been proposed. The primary histopathologic classification used for gastric carcinoma was first described in by Lauren [ ]. This system provided a general understanding of histogenesis and biology of this disease.
This classification simply divides gastric carcinomas morphologically into two types: diffuse and intestinal types, which have shown different genetic alterations and biologic behaviors. This classification has been applied to determine a clinical indication of endoscopic procedure or surgery and has supported unified epidemiologic data of gastric cancers by researchers. Intestinal type adenocarcinoma usually arises in the older population with an increased incidence in men and is frequently associated with chronic atrophic gastritis, intestinal metaplasia and Helicobacter pylori infection in neighboring mucosa [ ].
These tumors have a tendency to spread hematogeneously and often result in liver metastasis. Helicobacter pylori infection, high-salt diet and smoking have been recognized as risk factors of intestinal type adenocarcinoma [ 6 ]. In regard to the carcinogenesis of intestinal type gastric carcinoma, Correa et al proposed a multistep progression from Helicobacter pylori infection and gastritis to intestinal type gastric carcinoma [ ]. The sequence of changes in the stomach has been proposed that Helicobacter pylori infection or autoimmune gastritis causes atrophic gastritis with intestinal metaplasia, and transforms to dysplasia adenoma and further progress to adenocarcinoma.
Multiple genetic alterations or mutations occur and accumulate in each step of carcinogenesis and result in malignant transformation. However, this hypothetical model only can apply to intestinal type adenocarcinoma but not to diffuse type. Microscopically, intestinal type adenocarcinomas show well developed glandular structures, either with papillary or tubular component, surrounded by a variable degree of desmoplastic stroma and a mixed inflammatory infiltration Fig. In contrast, diffuse type adenocarcinoma frequently occurs in younger patients, with equal distribution among men and women [ ].
It tends to spread by direct tumor extension, resulting in peritoneal metastasis. Diffuse type adenocarcinoma has been believed to derive de novo from the peripheral stem cells of gastric gland neck proliferation zone without a recognizable precursor lesion [ , ]. Grossly, this tumor frequently shows ulcerations and occasionally combines with rigid, thickened, leather-bottle appearance of the gastric wall, called linitis plastica.
Microscopically, diffuse type adenocarcinoma is composed of individual tumor cells with or without signet ring cell configuration or small clusters of discohesive pleomorphic cells with little or no gland formation, which commonly deeply invade the full thickness layers of the gastric wall with desmoplastic reaction. Different clinicopathologic features of intestinal and diffuse type gastric carcinomas are listed in table 2. This disease is caused by germline mutation of CDH1 gene that encodes the cell adhesion protein E-cadherin, which plays an essential role in maintenance of the epithelial glandular structure [ ].
Diffuse gastric carcinoma is the main cause of cancer mortality in patients with CDH1 gene mutation [ ]. Some AGCs may not fit into one of two types clearly, and thus fall into mixed or unclassified categories. However, there have been only a few studies investigating the clinicopathologic features of mixed type AGCs. In recent series by Zheng et al, mixed type adenocarcinomas exhibit larger size, deeper invasion, more frequent local invasion, and lymph node metastasis compared to intestinal or diffuse type gastric carcinomas [ ].
Kozuki et al also demonstrated a similar result; prominent lymphatic permeation and lymph node metastasis were more frequently observed in mixed type than the pure type of gastric carcinoma [ ]. These findings suggest that mixed type adenocarcinoma of the stomach may have more aggressive behavior than two pure types and could be separated as a distinct entity. In , Ming proposed another classification system of gastric carcinoma based on the pattern of tumor growth and invasiveness as an indicator of biological behavior, expanding vs.
The expanding type adenocarcinomas grow predominantly by expansion with a sharply delineated periphery, resulting in a nodular growth of tumor. In contrast, the infiltrative type tumors show diffuse infiltration of tumor cells into the layers of gastric wall, without forming masses or nodules.
There are some overlapping features between Lauren and Ming classifications. In , Goseki et al proposed a classification system of AGC based on the degree of tubular differentiation and the amount of intracellular mucin production [ ]. This system divides AGCs into four groups based on tubular differentiation and mucin production by tumor cells. Four grades of tumor were proposed: group I: well differentiated tubules, intracellular mucin poor; group II: well differentiated tubules, intracellular mucin rich; group III: poorly differentiated tubules, intracellular mucin poor; group IV: poorly differentiated tubules, intracellular mucin rich.
Although the Lauren and other classifications provide a simplified categorization of usual gastric carcinomas and better understanding of their biology and behavior in large epidemiologic studies, they are less useful to apply to a variety of histologic subtypes of gastric carcinoma for predicting their clinical outcome. World Health Organization WHO proposed a classification to meet this need based on traditional histopathologic features and the degree of differentiation of gastric carcinoma [ ].
Gastric adenocarcinomas are graded like other glandular neoplasms based on the degree of glandular differentiation into well, moderately, and poorly differentiated subtypes. The degree of differentiation is considered as an important prognostic factor that is highly associated with the depth of tumor invasion and a risk of lymph node metastasis [ - ]. WHO classification has been also used for determining the therapeutic options of patients with gastric carcinoma. The degree of differentiation is one of the important criteria for performing endoscopic resection. In WHO classification, gastric carcinomas are divided into five categories based on histopathologic features including adenocarcinoma, adenosquamous carcinoma, squamous cell carcinoma, undifferentiated carcinoma, and unclassified carcinoma [ ].
Adenocarcinomas are subdivided into papillary, tubular, mucinous and signet ring cell types. Generally, papillary and tubular variant are classified into intestinal, expanding, or differentiated type, whereas, mucinous and signet ring cell variants are categorized into diffuse, infiltrative, or undifferentiated type [ ].
Tubular adenocarcinoma consists of tubular-shaped branching glands lined by pseudostratified columnar or cuboidal epithelium with elongated hyperchromatic nuclei having coarse chromatin and occasional mitotic figures Fig. Acinar structure may be present. The degree of cytologic atypia varies from low grade to high grade tumors. If tubular adenocarcinoma combines with papillary adenocarcinoma component, it is termed as a tubulopapillary variant. Gastric biopsy shows a well differentiated tubular adenocarcinoma. Adjacent to the carcinoma, regenerative foveolar epitheliums are admixed.
In high power examination, the anastomosing glands are composed of atypical cells with vesicular nuclei and prominent nucleoli. Tubular adenocarcinoma, well differentiated, in the resected stomach. Grossly, the tumor in the high body of greater curvature is advanced gastric carcinoma mimicking EGC type IIb.
In the superficial area, the carcinoma mimics regenerative changes due to its bland morphology. However, the carcinoma infiltrates into subserosa. Papillary adenocarcinoma is characterized by elongated finger-like processes that have a fibrovascular connective tissue core in the center and lined by cylindrical or cuboidal cells. The nuclear cytologic atypia varies from low to high grade.
Papillary adenocarcinomas have distinct clinicopathologic features such as a higher frequency in aged patients, proximal location, and elevated macroscopic type [ ]. Although papillary adenocarcinoma has been categorized into differentiated-type adenocarcinoma with low grade malignancy, some studies have shown that papillary adenocarcinomas of the stomach have a higher frequency of lymph node metastasis, liver metastasis and poorer surgical outcome compared to other types of gastric carcinoma [ - ].
Nakashima et al proposed a nuclear grading score for papillary adenocarcinoma of the stomach based on the extent of nuclear pleomorphism and nuclear polarity [ ]. They also reported that papillary adenocarcinoma with a high nuclear grade is usually accompanied by more advanced mural invasion, a higher risk of lymph node metastasis, higher chances of HER2 overexpression, and poorer prognosis.
This study suggested that papillary adenocarcinomas with high nuclear grade of the stomach may be a good therapeutic candidate for anti-HER2 trastuzumab therapy. Mucinous adenocarcinomas are also referred to as colloid, mucous, and muconodular carcinoma. They may present in one of two forms; tubular glands with mucus-secreting epithelium surrounding collections of extracellular mucin and signet ring cells floating in the mucinous lake [ ].
Sometimes, the tumor is predominantly composed of large acellular mucin pools with a few scattered tubular glands or signet ring cells Fig. The clinical outcome and prognosis of mucinous adenocarcinoma compared to non-mucinous adenocarcinoma is controversial. Some authors reported that the prognosis of patients with mucinous adenocarcinoma is poorer than that of patients with non-mucinous adenocarcinoma [ ]. However, others demonstrated that the 5-year survival rates of mucinous and non-mucinous adenocarcinomas are not different when the tumors are compared stage by stage [ ].
A recent study from South Korea reported that in mucinous gastric carcinomas, tumor size predicted prognosis more accurately than conventional pT stage depth of invasion in a study with large number of cases [ ]. Signet ring cell adenocarcinoma of the stomach is characterized by diffuse infiltration of signet-ring type of tumor cells into the gastric wall. Most signet ring cell carcinomas accompany with marked desmoplasia within tumor.
Signet ring cell carcinomas diffusely infiltrate through the muscular propria and subserosa with sparing the mucosa and present as firm and non-distensible texture, forming leather bottle appearance of stomach linitis plastica. In these cases, because a few scattered tumor cells are embedded in the desmoplastic stroma, it is easy to overlook the presence of malignant cells. Therefore, a careful pathologic examination is strongly recommended for delineating free margins and depth of tumor invasion during evaluation for frozen sections as well as permanent sections of gastrectomy specimens.
In problem cases, mucin stains periodic acid-Schiff, Alcian blue, mucicarmine and immunohistochemical staining for cytokeratin would be greatly helpful to demonstrate tumor cells. Li et al reported that the mean tumor size and depth of invasion of signet ring cell carcinoma is slightly larger and deeper than those of non-signet ring cell carcinoma [ ]. This tumor is frequently discovered at an advanced stage such as, stage IIIb and IV, and shows a higher rate of lymph node metastasis and peritoneal dissemination [ , ], and is associated with the poorer prognosis than other types of adenocarcinoma.
Mucinous carcinoma. Small clusters or strands of pleomorphic tumor cells containing mucin are present within mucin pool. Signet ring cell carcinoma. Signet ring cells showing foamy or pale basophilic abundant cytoplasm and an eccentrically located nucleus infiltrate the lamina propria. Gastric carcinomas have various amounts of differentiated tumor cells that may express heterogeneous phenotypes of mucin. Mucins are high molecular weight glycoproteins with complexity and diversity that constitute the major component of the mucus layer within the gastric epithelium.
Up to date, twelve core proteins of human mucin have been described. MUC1 and MUC5AC are expressed in the superficial foveolar epithelium and mucous neck cells of both the antrum and corpus, whereas MUC6 is expressed in the pyloric glands of antrum and the mucous cells of the neck zone of the corpus [ - ]. In contrast, MUC2 is found in the Golgi region of foveolar cells in the antrum and predominantly express in the areas of intestinal metaplasia with vacuolar staining in goblet cells [ , ].
MUC5B is expressed only during a brief period of fetal life [ ]. The expression of MUC2 is closely correlated with mucinous carcinoma and cardia adenocarcinoma [ ]. Expression of MUC1 is less frequent in adenoma compared to associated carcinoma [ ].
In contrast, MUC2, an intestinal type mucin, was highly expressed in the adenomas, but either persisted or decreased after malignant transformation to adenocarcinomas. These findings suggest that MUC2 expression would be an early event, while MUC1 expression would be a late event in the carcinogenesis of the stomach [ ].
The pattern of mucin expression may help to understand the differentiation pathway of gastric carcinoma and to predict its biologic behavior. However, it is still controversial whether expressions of mucin in gastric carcinoma have a prognostic significance or not. Immunohistochemical staining of cytokeratin 7 and cytokeratin 20 show various expression in the stomach.
Cytokeratin 20 is usually positive for antral epithelium, while cytokeratin 7 highlights the columnar cells of the cardia. It has known that adenocarcinomas of GEJ are more likely expressed CK7 and distal gastric adenocarcinoma are likely express CK20 [ ]. Although the carcinogenesis of gastric carcinoma is not clear, a rapid progress in the molecular biology of cancer helps us to understand a complex process of malignant transformation of the gastric epithelium caused by the accumulation of aberrant genetic mutations.
Gastric cancer is a heterogeneous disease with multiple environmental etiologies and alternative pathways of carcinogenesis. Beyond mutations in TP53, alterations in other genes or pathways account for only small subsets of the disease [ ]. Recent studies using next-generation sequencing NGS have revealed an extensive repertoire of potential cancer-deriving genes in several cancer types.
In stomach, recent exome sequencing data with 22 AGCs showed that genes involved in chromatin modification to be commonly mutated. Recent high throughput mutation profiling showed similar results [ , ]. Molecular targeted therapies have significantly emerged as an effective treatment and improved clinical outcomes of many common malignancies, including breast, colorectal, and lung cancers.
Although studies for targeting agents of gastric cancer did not show promising results in the past decades, recently, trastuzumab has been approved by US Food and Drug Administration FDA and European Medicine Agency as a first-line therapy in Human epidermal growth factor receptor HER 2 positive metastatic gastric cancers and GEJ cancers based on the result of a landmark clinical trial, so called ToGA Trastuzumab for Gastric Cancer study. Currently, many other molecular targeted agents for AGCs are undergoing clinical trials, including vascular epithelial growth factor VEGF inhibitor, other HER family targeted agents, and etc.
In this chapter, we focus on two main targeting agents, HER2 and VEGF inhibitors, and discuss about their biologic pathways and the results of clinical trials. The graph highlights the most significantly mutated genes of gastric cancer from the Cancer Gene Census. ARIDI1A seems to act like a tumor suppressor gene that involves DNA repair, differentiation, development, and has a regulatory role in proliferation [ , ].
ARIDI1A gene mutation has been reported in ovarian cancer, predominantly clear and endometrioid subtypes, and its rearrangement or deletion were identified in breast and lung cancer cell lines [ - ]. Recently, Wang et al demonstrated that gastric cancers with ARIDI1A mutation and loss are a distinct molecular subtype affecting predominantly EBV-positive and MSI-high gastric cancers, with a better prognosis and different carcinogenesis compared to the conventional gastric adenocarcinoma.
The MET proto-oncogene encodes a tyrosine kinase receptor for the hepatocyte growth factor and stimulates mitogenesis, motogenesis, vasculogenesis, and morphogenesis of the cells [ ]. In series of MET gene activation, tumors with c-met protein tended to display increased invasiveness and poorly differentiated histology with poor prognosis on multivariate analysis [ , ]. Fibroblast growth factor receptors FGFRs consist of four different variants and their overexpressions are associated with a poorer prognosis in patients with gastric carcinoma [ , ]. More detailed description of ERBB2 is in the next section.
The TP53 gene located at 17p Mutation or loss of heterozygosity LOH of the TP53 gene has frequently been demonstrated in gastric adenocarcinoma. However, in diffuse type gastric carcinomas, TP53 gene mutations are significantly increased in advance stage, indicating the importance of TP53 mutations for tumor stage progression.
Deletions or mutations of the adenomatous polyposis coli APC gene located on chromosome 5q21 have been detected in up to one third of gastric carcinoma cases [ ]. However, the frequency of APC gene mutations by histological types has been controversial and the role of APC gene in the carcinogenesis of gastric carcinoma is not clear yet compared to colorectal counterpart. RUNX3 is a recently discovered tumor suppressor gene that is involved in gastric carcinogenesis. Phosphatase and tensin homologue PTEN is located on chromosome 10q PTEN dephosphorylates the second messenger phosphatidylinositol-3,4,5-triphosphate, the product of phosphatidylinositolkinase [ ].
Mina et al reported that PTEN deletion was found in 8 of gastric cancer cases 4. Interestingly, Esteva et al reported that PTEN loss in patients with HER2 positive metastatic breast cancer was significantly associated with a poor response to trastuzumab therapy and a shorter survival time, suggesting its pivotal role in trastuzumab resistance [ ]. The Wnt signaling pathway plays an essential role in embryonic development and a variety of processes including cell cycle regulation in differentiated cells.
Mutations in the genes encoding Wnt pathway components are associated with various malignancies including tumors of gastrointestinal tract, in particular gastric cancer. E-cadherin, one of the members of the cadherins, acts as an adhesive molecule and plays an important role in growth development and carcinogenesis. E-cadherin gene located at chromosome 16q22 can be inactivated by mutation, LOH, and hypermethylation [ ]. Because E-cadherin is a components of adhering junctions, the mutations of E-cadherin and related genes result in the dyscohesiveness of tumor cells in the morphology of diffuse type adenocarcinoma.
Decreased expression of E-cadherin is predominantly found in undifferentiated, diffuse type gastric carcinomas, particularly in signet ring cell carcinoma and associated with invasiveness and a higher metastatic potential of gastric carcinomas resulting in a poorer prognosis [ , ]. Hereditary diffuse gastric cancer is caused by germline mutations of E-cadherin CDH1 [ ]. The Wnt signaling pathway would be less implicated for malignant transformation of intestinal type adenocarcinoma than diffuse type. CpG islands are DNA segments that are at least 0. Promotor CpG island hypermethylation is found in almost all cancers and involves in carcinogenesis and aging process by affecting on the tumor related genes and the inactivation of tumor suppressor genes.
Kang et al suggested that promotor CpG island hypermethylation occurs in an early step of the gastric carcinogenesis and accumulate during malignant transformation [ ]. MSI-high gastric carcinomas have some distinct clinicopathologic characteristic such as an association with intestinal type, distal stomach antrum , and more favorable prognosis compared to MSS and MSI-low carcinomas [ ]. In addition, some studies have shown that MSI-high gastric carcinomas have a lower risk of lymph node metastasis, near-diploid DNA content and tumoral lymphoid infiltration [ - ].
It is controversial that MSI involves in the early or late stage of gastric carcinogenesis due to contrary data. Particularly, EBV-positive gastric carcinoma is well known for global and nonrandom CpG island methylation of the promoter regions of cancer-related genes. Previous studies demonstrated that EBV-positive gastric carcinomas were strongly associated with CpG-island methylator phenotype and having multiple methylation of cancer related genes including genes of DNA repair and protection MLH1, MGMT, and GSTP1 , cell cycle regulation p14,15,16, and cox2 , cell adherence and metastasis E-cadherin, bcl-2, and p73 [ - ].
Recently, Park et al reported that in multiple gastric carcinomas, EBV infection allows the gastric mucosa to escape from aberrant methylation of MLH1 and induces a malignant pathway independent of MSI [ ]. Human telomerase reverse transcriptase hTERT is an important determinant of telomerase activity, the enzyme that catalyses the telomere DNA synthesis [ ]. It has been reported that the majority of intestinal type gastric carcinomas have shortened telomere length, high levels of telomerase activity and a significant expression of hTERT [ , ].
Therapeutic response and prognosis may be highly variable in patients with AGC within the same stage and chemotherapy regimens. Considering cancer is a product of accumulated genetic aberrations, elucidation of complex biological mechanism of cancer results in developing new molecular markers that would be specific for only tumor cells. Molecules that are closely associated with cell proliferation, invasion, and metastasis have been studied as potential candidates for targeted therapy.
HER2 c-erbB2 protooncogene is located on chromosome 17q The HER family proteins regulate cell growth, survival, adhesion, migration, and differentiation, which can be amplified or weakened in tumor cells. HER2 overexpression is strongly associated with poor clinical outcome and disease aggressiveness [ - ]. Recently, the ToGA trial which compared trastuzumab plus chemotherapy vs. It has been actively investigated as a candidate agent for patients with trastuzumab-resistant gastric and GEJ cancers [ ].
VEGF is a signal protein produced by cells that plays a key role in angiogenesis within a tumor and increases microvascular permeability [ ]. It has been well known that tumors cannot grow beyond a certain limited size if it does not have an adequate blood supply. Tumors produce and secret VEGF and related receptors to enhance neovascularization. Karayiannakis et al reported that the serum VEGF concentration is strongly associated with metastasis and poor prognosis in patients of gastric and GEJ carcinomas [ ].
Further details about methods and data sources are provided in the eAppendix, eFigures, and eTables in the Supplement. Box 2 contains a list of the supplementary figures and tables. Additional information is available from the authors in Web Tables 1 through 3; the web addresses for these items are listed in Box 3. Hereinafter, citations to Web Tables are for those given in Box 3.
- Rare cancers: the greatest inequality in cancer research and oncology treatment!
- Rare cancers: the greatest inequality in cancer research and oncology treatment;
- Pancreatic cancer - Diagnosis and treatment - Mayo Clinic?
- Womens Songs from West Africa.
- 20 Irish Phrases in 13 Minutes: Learn Irish Quickly and Easily.
- Stomach Cancer - Jefferson University Hospitals!
- CALL FOR PAPERS.
Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 8. Figure 9. Figure Table 2. Web Table 1. Incidence absolute numbers and rates by sex, , ja. Web Table 2. Mortality absolute numbers and rates by sex, , ja. Web Table 3. DALYs absolute numbers and rates by sex, , ja. Methodological updates were made to the mortality to incidence MI ratio estimation, which are described in detail in the eAppendix in the Supplement. Major updates for the MI ratio predictions were out-of-sample validation of multiple model types and selection of 1 model per cancer based on the out-of-sample root-mean-squared error.
For GBD , a sociodemographic index SDI was developed, which is a summary indicator derived from measures of income per capita, educational attainment, and fertility. Detailed methods describing computation of the SDI are reported elsewhere. The composite SDI index is the mean of the 3 rescaled components. An SDI of 1. Locations were grouped into quintiles based on their SDI value in Quintile cutoffs were based on the distribution of geography-years from to with the exception of populations smaller than 1 million. For GBD , the full time series from to was estimated.
We focus here on changes over the last decade. The initial process in the burden of cancer estimation is the modeling of cancer mortality. Mortality data from vital registration systems, verbal autopsies, and other sources like disease surveillance records were processed and added to a cause-of-death database. Methods and data sources have been described in detail previously.
To maximize data availability and take advantage of cancer registry data in countries with scarce mortality data, incidence data from cancer registries were transformed to mortality estimates through the use of separately estimated MI ratios. Modeling of the MI ratios is described in detail in the eAppendix in the Supplement. In brief, the estimation followed a 3-step approach, the creation of logit random effect models, spatiotemporal smoothing, and Gaussian process regression. A final model was selected based on out-of-sample validation. Updated cancer registry data for GBD was obtained from the GBD collaborator network or downloaded from publically available sources.
All data sources used for MI ratio estimation, as well as those used for incidence data transformed to mortality estimates, are listed in the eAppendix and eTable 2 in the Supplement. The number of site-years used by source type and by cancer can be found in eTable 3 in the Supplement. All data sources were extracted at the most detailed cause- and age-specific level and mapped to the GBD cause list.
Although NMSC is the most common cancer in many populations, most cancer registries do not include NMSC, which necessitates different estimation methods from the cancers presented here. Individual cause mortality estimates from CODEm were constrained to fit independently modeled, all-cause mortality estimates using the tool CodCorrect. Final mortality estimates were transformed into incidence estimates using modeled MI ratios. Uncertainty from the mortality estimation and from the MI ratio estimation was propagated to the incidence estimates. Ten-year cancer prevalence was modeled by estimating cancer survival using an MI ratio—based scaling factor, which takes into account location, year, and sex see the eAppendix in the Supplement for details.
This factor was used to scale the incidence cohort between a theoretical best-case and a theoretical worst-case survival. The absolute survival estimates allowed calculation of year prevalence for each incidence cohort. Duration of the 4 prevalence phases by cancer can be found in eTable 13 in the Supplement.
After dividing total prevalence into these 3 sequelae, we attributed the remaining prevalence to the remission phase. To calculate YLDs, the prevalence for each sequela was multiplied with a disability weight. Additional disability was estimated for procedures and procedure-related morbidities associated with the treatment of breast, larynx, colorectal, bladder, and prostate cancer mastectomy, laryngectomy, stoma, urinary incontinence, and impotence under the assumption that these are major disabling sequelae after cancer treatment.
Login to your account
Disability weights used for the different sequelae as well as methods to determine disability prevalence for these cancer-related outcomes can be found in the eAppendix in the Supplement. The sum of the YLDs for each general sequela, as well as for procedure-related sequelae, represent the total YLDs for each cancer. We calculated 2 scenarios to analyze the contribution of population aging, population growth, and changes in the age-specific incidence rates on the absolute change of cancer incidence. In the first scenario, the age structure, sex structure, and the age-specific rates from were applied to the total population of the year The difference between the total number of cases in and the hypothetical scenario were attributed to population growth.
In the second hypothetical scenario, the age-specific rates from were applied to the age structure, sex structure, and population size of Differences between the second hypothetical scenario and the first hypothetical scenario were attributed to population aging. Differences between the total number of cases in and the second hypothetical scenario were attributed to changes in the age-specific rates. The GBD world population standard was used for the calculation of age-standardized weights.
In , there were Cancer caused At the global level, the odds of developing cancer during a lifetime age years differed between the sexes: they were 1 in 3 for men and 1 in 4 for women eTable 16 in the Supplement. These odds differ substantially among SDI categories. In the lowest SDI quintile, the odds of developing cancer for men aged between 0 and 79 years were 1 in 6, whereas in the highest SDI quintile, 1 in 2 men developed cancer.
For women, the odds of developing cancer was 1 in 5 in the lowest SDI quintile and 1 in 3 in the highest quintile. The most common causes of cancer deaths for men were TBL, liver, and stomach cancer with 1. For women in , the most common incident cancers were breast, colorectal, and TBL cancer, with 2. Figure 1 shows the pattern of cancer incidence and mortality by age group. Leukemia, other neoplasms, and brain and nervous system cancers were also the leading contributors to childhood cancer deaths Figure 1 B.
For adolescents and young adults age years the most common cancers at the global level were breast cancer, cervical cancer, and other neoplasms. The main causes of cancer deaths for this age group were leukemia, other neoplasms, and liver cancer. For the population older than 39 years, the cancers contributing the most incident cases were TBL, breast, prostate, and colorectal cancer, while the main contributors to cancer deaths in this age group were TBL, stomach, and colorectal cancer.
Between and , age-standardized incidence rates ASIRs for all cancers combined increased in of countries or territories Figure 2. In contrast, age-standardized death rates ASDR for all cancers combined decreased within that timeframe in of countries or territories, as shown in Figure 3 , which also shows that countries with an increase in ASDR were largely located on the African continent.
- Are You Supporting a Missionary or Moochionary?.
- The Case Files of Doyle & McCraken?
- Gastric Carcinoma: Morphologic Classifications and Molecular Changes;
- All About Small Bowel Cancer | OncoLink?
- Change Password.
- Login to your account!
- Stomach Cancer;
However, these decreases were not statistically significant. The top 10 cancers were ranked highest top number of incident cases Figure 4. Breast cancer was the most common cancer overall, with an estimated 2. The vast majority occurred in women, with 2. Breast cancer was the leading cause of cancer in all SDI quintiles except for the high and high-middle SDI quintiles where it was the second most common cancer Figure 4.
For women, breast cancer was the leading cause of death in Table 1. Breast cancer caused One in 14 women and 1 in men developed breast cancer between birth and age 79 years eTable 16 in the Supplement at the global level. For women, the odds of developing breast cancer were the highest in high SDI countries, with 1 in 9 women developing breast cancer, compared with the lowest odds of 1 in 20 women in middle SDI countries developing breast cancer between age 0 and 79 years. Breast cancer was the most common cancer for women in countries or territories and the most common cause of cancer deaths in women in countries or territories eFigures 10 and 12 in the Supplement.
Between and , breast cancer remained the fifth leading cause of global cancer YLLs, as shown in Figure 6. Tracheal, bronchus, and lung cancer caused Men were more likely to develop TBL cancer than women, with 1 in 18 men and 1 in 45 women developing TBL cancer between birth and age 79 years eTable 16 in the Supplement.
In low SDI countries the odds were substantially lower, with 1 in 70 men and 1 in women developing TBL cancer between birth and age 79 years. It was the most common cause of cancer deaths by absolute cases globally as well as in all SDI quintiles except for countries in the low SDI group, where TBL cancer ranked seventh Figure 5. Web Tables 1 and 2. Tracheal, bronchus, and lung cancer was the cause of the most incident cases for men in 38 countries and the most common cause for cancer deaths in countries or territories eFigures 9 and 11 in the Supplement.
For women, TBL cancer was the most common cause of cancer deaths in 20 countries and territories eFigure 12 in the Supplement. Figure 8 shows slightly decreasing ASIR at the global level for men and increasing trends for women between and This trend was much more pronounced for the high SDI quintile. In , there were 1. The odds of developing colon and rectum cancer before age 79 years at the global level was higher for men than for women 1 in 28 men, 1 in 43 women, eTable 16 in the Supplement.
The highest odds were in the high SDI quintile, with 1 in 14 men and 1 in 23 women developing colorectal cancer compared with 1 in 94 men and 1 in women in the low SDI quintile. Globally, and for high SDI countries, colon and rectum cancer ranked third for cancer incidence and second for cancer deaths in as shown in Figures 4 and 5. Colon and rectum cancer incidence ranked lowest in low SDI countries as the eighth most common cancer and was the sixth leading cause for cancer mortality. Colon and rectum cancer was the cancer with the highest incidence in for men in 6 countries eFigure 9 in the Supplement.
For women, colon and rectum cancer was the most common cause of cancer deaths in 5 countries eFigure 12 in the Supplement. Colon and rectum cancer has remained the fourth leading cause for cancer YLLs between and Figure 6. Prostate cancer caused 6. The odds of developing prostate cancer between ages 0 to 79 years was 1 in 14 at the global level and ranged from 1 in 47 men for low-middle SDI countries to 1 in 6 men in high SDI countries eTable 16 in the Supplement.
In , prostate cancer was the cancer with the highest incidence for men in countries or territories, and the leading cause of cancer deaths for men in 29 countries eFigures 9 and 11 in the Supplement. Thirty-four percent of this increase can be attributed to a change in the age-specific rates. Stomach cancer caused One in 27 men and 1 in 68 women develop stomach cancer before age 79 years. The highest odds for men occurred in middle SDI countries 1 in 25 , whereas the lowest occurred in low-middle SDI countries 1 in Globally and for high SDI countries, stomach cancer ranked fifth for cancer incidence and third for cancer deaths in Figures 4 and 5.
In high-middle, middle, low-middle, and low SDI countries, stomach cancer ranked third for incidence. For cancer mortality in high-middle, middle, and low SDI countries, stomach cancer ranked third. For low-middle SDI countries it ranked second for cancer mortality. Stomach cancer was highest in absolute incidence in for men in 26 countries and territories and was the leading cause of cancer deaths in 11 countries eFigures 9 and 11 in the Supplement. For women it was the leading cause of cancer deaths in 4 countries eFigure 12 in the Supplement. Liver cancer caused Liver cancer was more common in men, with 1 in 45 men developing liver cancer before age 79 years compared with 1 in women at the global level.
The highest odds of developing liver cancer was in middle SDI countries, with 1 in 38 men and 1 in 96 women developing liver cancer, whereas the lowest odds were seen in low-middle SDI countries, with 1 in 98 men and 1 in women developing liver cancer during their lifetime eTable 16 in the Supplement. Globally, liver cancer ranked sixth for cancer incidence and fourth for cancer deaths in , as shown in Figures 4 and 5.
In low SDI countries, it ranked fourth for cancer incidence and first for cancer mortality, whereas in middle and high-middle SDI countries it ranked fourth and sixth, respectively, for cancer incidence but second for cancer mortality. Liver cancer was the most commonly diagnosed cancer in for men in 11 countries eFigure 9 in the Supplement and the most common cause of cancer deaths in 40 countries eFigure 11 in the Supplement.
Liver cancer was the most commonly diagnosed cancer for women in Mongolia eFigure 10 in the Supplement in and the leading cause of cancer deaths for women in 5 countries in eFigure 12 in the Supplement. Liver cancer remained the second leading cause of cancer YLLs between and Figure 6. This global trend, however, masks an increase in low and high SDI countries since ASIRs have been increasing for low SDI countries since ; in high SDI countries, rates decreased until the early s for men and the late s for women and then increased.
Non-Hodgkin lymphoma caused 6. One in 78 men and 1 in women at the global level developed NHL between birth and age 79 years. Globally, for both sexes combined in , NHL ranked seventh for cancer incidence and 11th for cancer deaths Figures 4 and 5. Figure 13 shows the slight increase in ASIRs between and graphically with very similar trends for men and women and all SDI quintiles.
In , leukemia caused One in 87 men compared with 1 in women developed leukemia between ages 0 and 79 years at the global level. The highest odds were seen in the high SDI quintile, with 1 in 64 men and 1 in women developing leukemia. The lowest odds occurred in low SDI countries, with 1 in men and 1 in women developing leukemia eTable 16 in the Supplement.
Leukemia ranked eighth for cancer incidence and ninth for cancer deaths at the global level in Figures 4 and 5. Leukemia was ranked lowest in high-SDI countries at 13th place eighth for cancer deaths. Leukemia led incident cases in for men in 5 countries eFigure 9 in the Supplement. Bladder cancer caused 3. Bladder cancer was more common in men, with 1 in 59 men being diagnosed before age 79 years compared with 1 in women. The odds of developing bladder cancer during a lifetime were the highest in high-SDI countries 1 in 36 men and 1 in women and the lowest in low-SDI countries 1 in men and 1 in women eTable 16 in the Supplement.
Globally, bladder cancer ranked ninth for cancer incidence and 13th for cancer deaths in , as shown in Figures 4 and 5. It ranked the highest in high-SDI countries at position 8 11th for mortality. Bladder cancer was the most commonly diagnosed cancer in for men in Egypt eFigure 9 in the Supplement. Globally, it dropped from the 17th to the 18th leading cause of cancer YLLs between and Figure 6.
If population age structure and size had remained the same in as they were in , bladder cancer incidence would have been stable. Rates increased in the low and low-middle quintiles. One in 68 women developed cervical cancer between birth and age 79 years at the global level eTable 16 in the Supplement. The odds were the highest in low SDI countries, with 1 in 24 women developing cervical cancer, and the lowest in high SDI countries, where 1 in women developed cervical cancer during a lifetime.
In , cervical cancer was the most commonly diagnosed cancer for women in 11 countries eFigure 10 in the Supplement and the most common cause of cancer deaths for women in 50 countries eFigure 12 in the Supplement. Seventy-one percent of this change can be explained by an increase in age-specific incidence rates Table 2. Twenty-eight percent of the change can be explained by an increase in the age-specific incidence rates Table 2.
Countries with increasing cancer mortality rates were dominantly in Sub-Saharan Africa where, with few exceptions, the complex health care infrastructure required to treat cancer is generally lacking. Prevention and treatment of chronic hepatitis B and C, which account for the majority of liver cancer deaths, would reduce the incidence and mortality of liver cancer. However, rates for high SDI countries have increased since , and rates for low SDI countries increased in the most recent observations from to These findings are consistent with observations in some high-income countries, where obesity, diabetes, and hepatitis C are thought to be major contributors to rising incidence rates.
Human papillomavirus vaccination is universally recommended by health authorities and is expected to reduce cervical cancer incidence over the next decades if vaccination uptake is successful. In the meantime, screen and treat approaches that have been shown to reduce cervical cancer mortality in high-income countries should be implemented in regions with a high burden of cervical cancer. Stomach cancer rates have been declining globally for decades. In the highest SDI quintile, rates declined until but have since increased.
One possible explanation is the increasing trend of gastric cancer of the cardia in high-income countries owing to risk factors such as obesity, although the mechanism explaining this association is not fully understood. Although studies have shown reductions in stomach cancer mortality in screened populations, the target group, onset, and modality of screening programs remain controversial.
Prevention and treatment of carcinogenic infections that lead to these observed cancer patterns in low SDI countries have the potential to decrease future cancer burden as well as reducing associated diseases like cirrhosis in endemic areas. The dominance of infection-related cancers in low SDI countries is an exceptional pattern compared with the leading causes of cancer deaths for countries in higher SDI quintiles, where TBL, colorectal, stomach, liver, and breast cancer lead the rankings.
The importance of tobacco control as a crucial cancer control strategy should therefore not be underestimated, especially since tobacco control has health benefits reaching far beyond cancer prevention. Early detection strategies can range from improving breast cancer awareness and clinical breast examination in basic resource settings to mammography screening.
The variation in the leading causes of cancer incidence and mortality between countries documented in the GBD study are remarkable. The largest ranges in rankings for cancer mortality were found in cervical cancer, acute lymphoid leukemia, and nasopharynx cancer followed by esophageal, lip and oral cavity, gallbladder and biliary tract cancer, and melanoma eFigure 8 in the Supplement.
For cancer incidence, the largest divergence was seen in esophageal cancer, cervical cancer, melanoma, acute lymphoid leukemia, kidney, larynx, and gallbladder and biliary tract cancer eFigure 7 in the Supplement. For most of these cancers, the observed pattern can be explained by known risk factors or by access to care. Risk factors for esophageal cancer vary depending on the histologic characteristics of squamous cell carcinoma vs adenocarcinoma. For example, it is postulated that the diverse incidence of esophageal squamous cell carcinoma, the dominant histologic type in high-endemic areas, is driven by chronic cell damage from risk factors like smoking, alcohol consumption, heat damage, and nutritional deficiency.
A detailed analysis of the reasons behind the observed variation in cancer burden goes beyond the scope of this analysis. However, the cited examples show that the descriptive epidemiology approach of the GBD study can identify patterns of cancer burden that can be hypothesis generating. This is especially true when cancer burden is analyzed in conjunction with nonmalignant diseases that might share similar risk factors—a potential provided by the comprehensive nature of the GBD approach. As in prior GBD studies, our estimates depend on the quality and quantity of the data sources available to inform the estimates.
Because of the lag time for data reporting, estimates for are mainly based on data and trends from recent years. For many countries, data sources for informing cancer burden estimation are still sparse, and the GBD estimates rely heavily on covariate selection in the models and regional patterns.
Cancer registration has a long tradition in many countries. However, in regions where the burden of cancer is expected to grow significantly due to anticipated population growth and aging, cancer registries often only cover a small fraction of the population, are of low quality, or do not exist.
Cancer mortality estimates are predominantly based on vital registration data, cancer registry data, and to a much lesser extent other data sources. If a large proportion of deaths are miscoded, the redistribution of these so-called garbage codes can substantially affect mortality estimates. Since GBD cancer incidence estimates are based on mortality estimates, garbage code redistribution directly affects cancer incidence.
Misclassification of metastatic sites as primary cancer sites eg, liver, lung, brain is another source of potential bias, especially in locations with limited diagnostic resources. Changes in coding practices or coding systems may also have an effect, even though mapping to the GBD causes list includes adjustments to account for different coding systems. Cervical and uterine cancer incidence rates are potentially overestimated in the GBD in locations where hysterectomies are common, since rates are calculated without adjustments for the population at risk.
For GBD , we have updated the MI mortality to incidence ratios and used out-of-sample validation of a set of potential models to choose the most appropriate MI model.
Journal of Gastroenterology and Hepatology Research
However, in young age groups with sparse data, and in areas where no matching mortality and incidence data exist which is the case for most countries in Sub-Saharan Africa , the MI model is based on the combination of trends for older age groups as well as covariate selection. However, when interpreting the results, it is important to recognize that those data sources for the MI ratios that determine incidence for leukemia subtypes are mostly from high-income countries.
Changes in classification of leukemia subtypes over time can also have an effect on the GBD estimates. With the wider availability of cancer registry data, including childhood cancer registries, it is expected that estimates will continue to be adjusted in future iterations of the GBD. Despite significant reductions in cancer mortality in many countries, cancer poses a barrier to future development.
Incidence is expected to increase, straining resources even in countries with advanced health care systems. An expanding arsenal of cancer prevention and treatment interventions, together with a political commitment to address NCDs, offers hope that this threat can be controlled. The GBD study enables timely tracking of progress toward defined targets. Correction: This article was corrected on March 9, , to add additional contributions reported after publication.
Published Online: December 3, Johnson, PhD; Jost B. Lim, PhD; Alan D. Lopez, PhD; Michael F. Schwartz, PhD; Katya A.