Society: AGA

BACKGROUND
Early-onset colorectal cancer (CRC diagnosed before age 50) has risen worldwide, with an increasing number of survivors of reproductive age. We aimed to investigate the risk of adverse pregnancy and neonatal outcomes among early-onset CRC survivors.

METHODS
We conducted a nationwide study of 207 births in women with early-onset CRC and 1019 births in women without CRC from the general Swedish population (1992-2019), matched on age, calendar year, parity, and county of residence. To further adjust for confounding, we identified 146 births in female siblings of women with early-onset. Early-onset CRC cases were identified through the Swedish Cancer Register. Sibling identification and outcome data were retrieved through linkage of the Swedish Multi-generation Register, Medical Birth Register, and National Patient Register. Using conditional logistic regression, we estimated multivariable-adjusted odds ratios (ORs) and 95% confidence intervals (CIs).

RESULTS
Compared to women without prior CRC, early-onset CRC survivors who gave birth had increased risk of pre-eclampsia (7.2% vs 3.2%; OR=2.52, 95%CI 1.25-5.08), any Cesarean delivery (C-section) (24.6% vs 19.4%; OR=1.43, 95%CI 1.00-2.06), particularly emergency C-section (17.4% vs 10.5%; OR=1.79, 95%CI 1.17-2.75), after adjustment for maternal education level, country of birth, body mass index and smoking in early pregnancy, and comorbidities. Sibling analyses showed similar trends. Maternal history of early-onset CRC was also associated with offspring preterm birth (12.1% vs 5.2%; OR=2.31, 95%CI 1.34-3.99), delineated as spontaneous (OR=1.06, 95%CI 0.47-2.39) or medically-indicated preterm birth (OR=4.48, 95%CI 2.05-9.79). There was no increased risk of congenital malformation or small for gestational age birth.

CONCLUSIONS
In this population-based study, maternal history of early-onset CRC was associated with risk of both adverse pregnancy (pre-eclampsia, C-section) and neonatal outcomes (preterm birth).

BACKGROUND & AIMS
Substantial geographic variability in gastrointestinal (GI) cancer mortality has been reported in the United States. There is a substantial variation in the socioeconomic, ethnic composition, and social infrastructure among US counties. Whether underlying variation in social vulnerabilities among counties influence GI cancer mortality is uncertain. The aim of this study was to examine the association between county-level social vulnerability and GI cancer mortality.

METHODS
In this cross-sectional study in the US (2014-2020), we linked county-level Social Vulnerability Index (SVI) with county-level age-adjusted GI cancer mortality. SVI comprised of four subcomponents (socioeconomic status; household composition and disability; minority status and language; and housing type and transportation) and was created using 15 social attributes. We categorized SVI into quintiles based on the distribution among US counties (1^st [least vulnerable] to 5^th [most vulnerable]). We used generalized estimating equation to estimate the age-adjusted GI cancer mortality and rate ratio (RR) by quintiles of SVI. We examined mortality for all GI cancers and five individual GI cancers (esophageal cancer, gastric cancer, liver cancer, pancreatic cancer, and colorectal cancer). We adjusted for lifestyle factors in multivariable models including obesity, smoking, physical activity, alcohol consumption, and diet.

RESULTS
There were 1,101,223 GI cancer deaths from 2014 to 2020 among a population of 2,267,748,585. The largest concentration of counties with more social vulnerabilities and higher GI cancer mortality were clustered across the southwestern and southeastern parts of the US (Figure). The age-adjusted mortality rates per 100,000 population (95% CI) for all GI cancers, esophageal cancer, gastric cancer, liver cancer, pancreatic cancer, and colorectal cancer were 42.5 (42.0-43.0), 4.3 (4.1-4.4), 3.0 (2.9-3.1), 7.1 (6.9-7.2), 11.6 (11.4-11.7), and 15.3 (15.0-15.6) (Table). There was an increase in overall GI cancer mortality with higher SVI (RR_{Q5 vs Q1}, 1.22; 95% CI, 1.19-1.26). This was also observed in all individual GI cancer types (RR_{Q5 vs Q1} ranging from 1.08 to 1.63). After adjusting for lifestyle factors, only gastric cancer (RR_{Q5 vs Q1}, 1.46; 95% CI, 1.33-1.59) and liver cancer mortality (RR_{Q5 vs Q1}, 1.31; 95% CI, 1.20-1.43) remained significantly associated with SVI.

CONCLUSIONS
In this analysis, US counties with more social vulnerabilities had higher GI cancer mortality, some of which could be attributed to disparities in lifestyle risk factors. Focused public health interventions should be directed to counties with higher social vulnerability to curb the growing burden of GI cancer.

<b>Figure. Social Vulnerability Index (2014–2018) and GI cancer mortality rates (2014–2020) in the US. </b>(A) Counties by Social Vulnerability Index; (B) by age-adjusted GI cancer mortality rates. Q indicates quintile.

Figure. Social Vulnerability Index (2014–2018) and GI cancer mortality rates (2014–2020) in the US. (A) Counties by Social Vulnerability Index; (B) by age-adjusted GI cancer mortality rates. Q indicates quintile.

<b>Table. Association between Social Vulnerability Index and age-adjusted mortality rates for GI cancer, 2014-2020</b>

Table. Association between Social Vulnerability Index and age-adjusted mortality rates for GI cancer, 2014-2020

Introduction: Incidence rates of colorectal cancer (CRC) are increasing in younger (age 18-49 years) and middle-age (age 50-59 years) adults in the U.S. Rates have increased by birth cohort, a phenomenon that informed recommendations to lower the age to initiate average-risk screening. This birth cohort effect has not yet been examined in other parts of the world. To address this gap, we conducted an age-period-cohort analysis to test for birth cohort effects in seven world regions.

Methods: We used data from the Global Health Data Exchange to identify adults (age 20-89 years) newly diagnosed with CRC during 1990 – 2019 in seven world regions: East Asia, Europe and Central Asia, Latin America and the Caribbean, the Middle East and North Africa, North America, South Asia, and Sub-Saharan Africa. The Global Health Data Exchange comprises 204 countries and territories and uses data from vital registration, verbal autopsy, and population-based cancer registries to generate estimates of incident cancers. We used five-year age groups (e.g., 20-24) and time periods (e.g., 1990-94) to create 17 birth cohorts (1910-14 to 1990-94). Separately by region, we estimated the ratio of age-specific incidence rates of CRC in each birth cohort relative to the 1950-54 birth cohort. We report incidence rate ratios (IRR) and 95% confidence intervals (CI).

Results: Age-specific incidence rates of CRC increased across birth cohorts after 1950-54 in all regions of the world, with the exception of Europe and Central Asia. Compared to persons born in 1950-54, incidence rates of CRC were up to two times as high among persons born in the 1980s and 90s (Image). This pattern was more prominent in East Asia, Latin America and the Caribbean, and the Middle East and North Africa, but less so in Sub-Saharan Africa. For example, in East Asia, IRRs were 1.98 (95% CI 1.85, 2.13) and 2.53 (95% CI 2.23, 2.87) for the 1980-84 and 1990-94 birth cohorts, respectively. There were small increases for persons born in or after 1955-59 in Sub-Saharan Africa, with IRRs ranging from 1.03 (95% CI 1.02, 1.04) for the 1955-59 birth cohort to 1.16 (95% CI 1.11, 1.21) for the 1990-94 birth cohort. There was no increase by birth cohort in Europe and Central Asia.

Discussion: In nearly all regions of the word, incidence rates of CRC have increased among persons born after 1950-54, with particularly notable increases in East Asia, Latin America and the Caribbean, and the Middle East and North Africa. These increases occurred despite differences in the age structure, screening programs, diet and lifestyle, and diagnostic factors among world regions. Identifying environmental exposures introduced in and increasingly prevalent after 1950 (e.g., pesticides, flame retardants) may hold the key to finding risk factors responsible.

Background: Upper gastrointestinal (UGI) cancer is the second leading cause of cancer-related mortality in the US, and is increasing in incidence. Recent studies have shown variable incidence patterns based on subsite, but the trends of different histologic classes are not known. This study aims to evaluate trends in UGI adenocarcinoma vs non-adenocarcinoma cancers, stratified by age and sex, in a nationwide population-based cohort.

Methods: Patients diagnosed with esophageal, gastric, duodenal, or pancreatic cancer in the US from 2000-2019 were obtained from the SEER database (biliary, gallbladder, and liver cancer were excluded because they overwhelmingly have only one histologic classification). The data was accessed using SEER*STAT program (v 8.4.0.1, NCI), and cases were identified using ICD-O-3 codes of the above sites. The primary outcomes were incidence rates of adenocarcinoma and non-adenocarcinoma (calculated per 100,000, age-adjusted to the year 2000 US population), stratified by sex and age (< 55 years for young adults, and ≥55 years for older adults). Non-adenocarcinoma histology includes squamous cell carcinoma (SCC), neuroendocrine tumor (NET), gastrointestinal stromal tumor (GIST), and lymphoma, and each histology class was analyzed separately. Trends were assessed using the Joinpoint Regression Program (v4.9.1.0, NCI), and annual percentage change (APC) and average APC (AAPC) were calculated. Pairwise comparison was performed to assess identicalness and parallelism.

Results: A total of 697,305 patients with UGI cancer were identified (Table 1). The majority of cases were male (59.4%), ≥55 years of age (84.7%), and had adenocarcinoma (79.9%). Among the non-adenocarcinoma cases, the most common histological classes were NET (35.1%) and SCC (33.3%). Adenocarcinoma occurred in relatively greater proportion of men compared to NET or GIST (60.5% vs 49.9% and 50%, respectively; p < 0.001).

The incidence of UGI adenocarcinoma was relatively unchanged in men (APC 0.07%, p = 0.22) and slightly increased in women (0.46%, p < 0.001), and the overall number of SCC and lymphoma cases decreased (Fig.1). In contrast, the number of NET and GIST cases increased (APC 0.79%, p < 0.001). The rate of increase was higher in women (AAPC 1.13%; AAPC difference 0.72%, p < 0.001). Additionally, cases of GIST and lymphoma increased disproportionately in young women (AAPC difference 2.95% and 2.62%, respectively; p < 0.001), and was non-parallel and non-identical (p < 0.001).

Conclusion: The increased incidence of UGI cancer appears to be primarily driven by NET and GIST cases. The rate of increase in these cases is 2.8-fold higher in women compared to men. These findings suggest a sex-based disproportional exposure that may contribute to this rapid rise. Further studies are needed to elucidate possible causes of these disparate trends.

Table 1: Trends in Incidence of Upper GI adenocarcinoma vs non-adenocarcinoma, stratified by age and sex

Table 1: Trends in Incidence of Upper GI adenocarcinoma vs non-adenocarcinoma, stratified by age and sex

Fig.1 Trends in Upper GI Adenocarcinoma vs Non-adenocarcinoma, Stratified by Age, Sex, and Primary Site

Fig.1 Trends in Upper GI Adenocarcinoma vs Non-adenocarcinoma, Stratified by Age, Sex, and Primary Site

Background Colorectal cancer (CRC) is the third most diagnosed and the second most common cause of cancer death worldwide. The incidence of CRC in the US is declining in adults 50 years and older; however, recent studies suggest an increasing disease burden among US adults under age 50. This study aims to compare the incidence, mortality, and mortality-to-incidence ratios (MIRs) of CRC in EU15+ countries versus global rates to determine if similar age-stratified occurrences are observed across these developed EU15+ countries with similar western culture-related risk factors.

Methods We conducted an observational cross-sectional study for CRC globally and in EU15+ countries between 1990 and 2019 using the Global Burden of Disease (GBD) database. Incidence and mortality rates were extracted, and MIRs were calculated. All rates were reported per 100,000 population. GBD utilizes C18-C21, D01.0-D01.3*, D37.3-D37.5* ICD-10 codes and 153-154, 230.3-230.6* ICD-9 codes for colorectal cancer. The data were age-stratified into groups of adults between ages 25-49, 50-69, and 70+ years.

Results The incidence of CRC increased globally for all age groups from 1990 to 2019, with the highest increase observed for males (+75.9%) and females (+27.7%) aged 25-49. A similar trend was observed in 15/19 EU15+ countries for males and 16/19 for females aged 25-49. Incidence rates decreased in 18/19 EU15+ countries in males 50-69 and in 10/19 EU15+ countries in females 50-69. For males 70+, incidence rates decreased in 18/19 EU15+ countries, whereas in females 70+, incidence rates decreased in 6/19 countries. Global mortality rates decreased for all age groups in females while increasing for males in all age groups. Mortality rates for EU15+ countries overall exhibited a reduction in mortality, with a few exceptions depending on age and gender. Despite increased incidence rates, MIRs in all subpopulations declined globally and in all subgroups of EU15+ countries except for 7/19 countries for females 70+.

Conclusions GBD data demonstrates an increase in the incidence of CRC in men and women between 25-49 years from 1990 to 2019, both globally and among EU15+ countries. In comparison, the incidence of CRC in men and women over 50 increased globally but declined in a majority of EU15+ countries. We believe this represents both the success of current CRC screening efforts in the western world (historically directed at age 50+), as well as an alarming trend toward higher CRC burden in adults under age 50. The latter may be due to modifiable risk factors but does not appear unique to the EU15+ countries. Fortunately, the decline in MIR that we observed in all subpopulations both globally and in EU15+ countries likely reflects improvement in cancer care since 1990.

<b>Figure 1: </b>Incidence rates of EU15+ countries and global rates for males and females in age groups 25-49, 50-69, and 70+ years.

Figure 1: Incidence rates of EU15+ countries and global rates for males and females in age groups 25-49, 50-69, and 70+ years.

<b>Figure 2: </b>Mortality rates of EU15+ countries and global rates for males and females in age groups 25-49, 50-69, and 70+ years.

Figure 2: Mortality rates of EU15+ countries and global rates for males and females in age groups 25-49, 50-69, and 70+ years.

Background
Population-based colorectal cancer (CRC) screening program by using faecal immunochemical test (FIT) can reduce CRC risk by identifying subjects with adenoma followed by polypectomy during confirmatory colonoscopy. Surveillance colonoscopy at regular intervals for these high-risk subjects is recommended. Current recommendations on surveillance intervals are devised by using the number, size, and histological findings of adenoma to reduce their CRC risk to the level of the average risk population. Given the potential of FIT in predicting CRC risk, this study aimed to develop a personalized surveillance interval for this population guided by faecal hemoglobin concentration (fHbC, μg Hb/ g faeces) level measured by FIT.
Materials and Methods
A prospective cohort study was applied to the subjects with colorectal adenoma detected in Taiwan National Colorectal Cancer Screening Program and received between 2010 and 2015. Taiwanese aged between 50 and 74 without colorectal cancer diagnosed are provided with biennial FIT-based screening. Subjects with fHbC higher than 100 are referred for confirmatory colonoscopy and treatments. The development of CRC for this cohort up to the end of 2018 was traced by the linkage to the Taiwan Cancer Registry. The CRC risk following polypectomy stratified by fHbC level measured at detecting adenoma and the size and histological type of the adenoma was used to gauge the surveillance interval.
Results
A total of 89771 subjects including 45377 (50.5%) with low-risk adenoma (large than 10 mm or with villous component) and 44394 (49.5%) with high risk-risk adenoma (smaller than 10 mm and without villous component) were enrolled in this study. During the average follow-up period of 5.5 years, 1413 CRC were identified, corresponding to the average risk of 2.8 (95% CI: 2.7-3.0 per 1000). An increasing trend of CRC risk by fHbC level ranging from 2.2 (20-49) to 4.0 (>=450) per 1000 was observed. By using the population average CRC risk of 10 per 1000, surveillance intervals of 3 and 5 years are required for subjects with high-risk and low-risk lesions, respectively (Figure 1). Among the low-risk lesions, the surveillance intervals of 6 to 3 years can be applied for the fHbC level of 20-49 to >=150 (Figure 2 (a)). The 3-year surveillance interval for subjects with high-risk adenoma was generally feasible, although a shorter interval (2.5 years) may be required for those with a high fHbC of >=150 (Figure 2 (b)).
Conclusions
The suggested interval of surveillance colonoscopy for subjects with high-risk adenoma was 3 years regardless of fHbC, although a shorter interval of 2.5 years may be considered for those with an fHbC level higher than 150. For subjects with low-risk adenoma, the intervals of surveillance colonoscopy ranging between 3 (>=150) and 6 years (20-49) stratified by the level of fHbC can be applied.

Figure 1

Figure 2

Han-Mo Chiu

National Taiwan University Hospital

Appearances