Cancer Epidemiology Biomarkers & Prevention最新文献

Background: Identifying changes in geographic disparities of cancer mortality reveals locations where cancer prevention and control efforts should be focused/targeted. We use recent cancer surveillance data to demonstrate the geographic disparity of major cancer mortality rates in the United States and its shift compared with previous data.

Methods: This cross-sectional study used the 2018 to 2022 county-level mortality rates of colorectal, lung, breast, and prostate cancers from the Centers for Disease Control mortality data. Counties with suppressed death counts were imputed by spatial regression models. Getis-Ord Gi* statistics were used to evaluate the spatial clustering of county mortality. Identified hotspot counties were visualized and compared with literature for hotspot pattern change.

Results: A total of 3,108 US mainland counties were included. Cancer mortality rates were significantly higher in 244 counties for colorectal, 456 for lung, 147 for breast, and 180 for prostate cancers. Hotspot areas were central Appalachia (colorectal and lung cancers), Lower Mississippi Delta (colorectal, breast, and prostate cancers), Midwest (colorectal and lung cancers), north Michigan/Wisconsin (lung and prostate cancers), north Florida (lung cancer), and the West (prostate cancer).

Conclusions: West central Appalachia and Lower Mississippi Delta continue to be hotspots for major cancer types, whereas previously identified eastern North Carolina/Virginia hotspots shrunk, east Oklahoma and North Florida emerged as new hotspots for lung cancer, and several hotspots emerged in the West for prostate cancer.

Impact: This study updated the analyses for geospatial disparity in major cancer mortality since 2018, illustrating recent changes in the disparity pattern and pinpointing areas that cancer prevention and control efforts should target.

Background: Colorectal cancer is the second leading cause of cancer death in Asian Americans. Asian Americans are a diverse, heterogeneous population composed of groups with differing cancer risk factors. Few prior studies have analyzed colorectal cancer mortality by disaggregated Asian racial subgroup.

Methods: Using 2005 to 2020 US national mortality records linked to American Community Survey one-year population estimates, we report age-standardized mortality rates per 100,000 person-years, standardized mortality ratios (SMR), and average annual percent change trends for the six largest Asian subgroups in a serial, cross-sectional study design. We compared these rates with non-Hispanic Whites. We stratified rates by sex, nativity, and colorectal cancer location (colon vs. rectum).

Results: Asian subgroups demonstrated substantial heterogeneity in colorectal cancer mortality. Relative to the non-Hispanic White group, Asian Indian Americans had the lowest rate [female SMR = 0.3; 95% confidence interval (CI), 0.3-0.3 and male SMR = 0.3; 95% CI, 0.3-0.3] and Japanese Americans the highest rate (female SMR = 0.9; 95% CI, 0.8-0.9 and male SMR = 0.9; 95% CI, 0.9-1.0). Chinese, Filipino, Korean, and Vietnamese Americans demonstrated mortality between Asian Indian and Japanese. Over the study period, most Asian subgroups had stable or decreasing mortality. However, both Korean and Vietnamese colorectal cancer mortality increased over the period. By the end of the study period, Korean Americans had the highest colorectal cancer mortality of any Asian subgroup.

Conclusions: Asian subgroups demonstrate heterogeneity in patterns of colorectal cancer mortality, emphasizing the necessity of disaggregation in cancer research.

Impact: Our study provides disaggregated Asian subgroup colorectal cancer mortality data, which may allow for targeted risk attenuation efforts.

Background: Along with incidence and mortality, temporal trends of cancer survival are a crucial part of cancer surveillance and control. The most common reported statistic is net survival, usually age standardized to an external reference population. However, net survival has an awkward interpretation, which has led to confusion and misunderstanding.

Methods: We describe the use of reference-adjusted all-cause survival, and the crude probability of death as an alternative to net survival for the analysis of temporal trends in cancer survival. Reference-adjusted measures aim to enable fair comparisons by incorporating additional reference-expected mortality rates into the estimation process. The different approaches are illustrated using data on 95,285 women diagnosed with breast cancer in Norway from 1986 to 2021.

Results: We compare different age distributions for age standardization and describe how using a recent calendar period for both the reference-expected mortality rates and age distribution for standardization leads to simple interpretation.

Conclusions: Reference-adjusted measures for monitoring temporal trends in cancer survival can lead to improved understanding and is of more relevance to patients and policy makers who live and make decisions in the real world. Using the most recent calendar period for both the age standard and the reference-expected mortality rates leads to simple and useful interpretation of the measures.

Impact: Increasing the use of reference-adjusted measures in the analysis of population-based cancer studies will enhance the understanding of cancer survival trends. The freely available software increases the likelihood of uptake.

Background: The use of asbestos-containing products was banned in Australia in 2003. However, the rates of new cases of mesothelioma, which has a very long latent period between exposure and disease, have continued to increase. The aim of this study was to investigate mesothelioma incidence in Australia by year of birth and age-period-cohort analysis and to develop projections of expected mesothelioma cases until 2034.

Methods: Data were derived from the Australian Cancer Database which provides complete national records of mesothelioma cases notified between 1982 and 2020. Incidence rates were age-standardized to the 2001 Australian standard population to enable comparisons of the population across time. Age-period-cohort models were used to examine the temporal trends of incidence rates by age, calendar year, and birth cohort. Projections for incidence rates of mesothelioma for 2020 to 2034 were estimated using Nordpred models.

Results: Graphs of age-standardized incidence rates of mesothelioma suggest a birth cohort effect, and the age-period-cohort model confirmed this. There was a birth cohort effect in all cohorts born before 1960, strongest in cohorts born during 1920 to 1949. Projection modeling to 2034 suggested that the age-standardized rates will continue to decline whereas crude incidence rates of mesothelioma will stabilize and then gradually decline, mostly among people of 60 to 84 years of age.

Conclusions: The findings are consistent with the greatest risk of mesothelioma in Australia occurring in cohorts with the highest levels of historical cumulative occupational exposure, showing the value of a ban on asbestos.

Impact: The number of new cases of mesothelioma per year is not expected to decline until after 2030.

Background: To compare cost-effectiveness of three novel noninvasive tests [multitarget stool RNA (mt-sRNA), multitarget stool DNA 2.0, and cell-free DNA] with guideline-recommended tests for colorectal cancer screening from payer's perspective.

Methods: Outcomes of a hypothetical cohort of 100,000 individuals aged 45 years with average colorectal cancer risk (no prior colorectal cancer diagnosis, adenomatous polyps, or other disorders associated with a high lifetime risk of colorectal cancer) in the United States were simulated by a lifelong Markov model. Screening strategies included guideline-recommended strategies (colonoscopy, flexible sigmoidoscopy, computed tomographic colonography, fecal immunochemical test, high-sensitivity guaiac-based fecal occult blood testing, and multitarget stool DNA), three novel noninvasive tests, and no screening. Scenario analyses on perfect (100%) and test-specific adherence (reported real-world adherence) were conducted. Outcomes included direct cost, quality-adjusted life-years (QALY), and incremental cost-effectiveness ratios (ICER).

Results: All screening strategies (vs. no screening) reduced colorectal cancer cases and deaths. In the perfect adherence scenario, every-10-year colonoscopy was the preferred strategy (ICER = US$261/QALY). In the test-specific adherence scenario, every-3-year mt-sRNA was the preferred cost-effective strategy (ICER = US $95,250/QALY). Testing cost, performance, adherence, and colorectal cancer prevalence, progression rate, and utility were influential factors. Every-3-year mt-sRNA showed the highest probability (37.6%) to be cost-effective in the test-specific adherence scenario at a willingness to pay US $100,000/QALY.

Conclusions: All strategies were cost-effective compared with no screening. Every-3-year mt-sRNA (preferred strategy in the real-world adherence scenario) provides a cost-effective alternative when adherence to colorectal cancer screening or follow-up was not perfect in clinical practice.

Impact: This is the first study to demonstrate cost-effectiveness of novel noninvasive tests versus all guideline-recommended colorectal cancer screening tests. See related In the Spotlight, p. 1053.

Background: Beneficial effects of cruciferous vegetable intake on breast cancer survival have long been postulated because they are primary sources of isothiocyanates, phytochemicals with multifaceted anticancer activities. However, observational studies have reported inconsistent results. We hypothesized that variations in vegetable types and polymorphisms in isothiocyanate-metabolizing genes across self-identified race and ethnicity contribute to such inconsistencies.

Methods: In the Pathways Study, a prospective cohort study of women diagnosed with breast cancer between 2005 and 2013 at Kaiser Permanente Northern California, cruciferous vegetable intake was assessed at diagnosis using food frequency questionnaires. Functional polymorphisms in isothiocyanate-metabolizing genes were identified in the literature and genotyped. Cox proportional hazards models were used to calculate hazard ratios (HR) and 95% confidence intervals (CI). The analysis included 3,656 (2,489 non-Hispanic White, 241 Black, 463 Asian, 378 Hispanic, and 85 others) participants.

Results: An overall inverse association between cruciferous vegetable intake and risk of total invasive events, including recurrence, second primary cancers, and death, was observed in age-adjusted models (HR, 0.86; 95% CI per serving, 0.77-0.97), whereas no significant dose-dependent associations were observed in multivariable analyses (HR, 0.91; 95% CI per serving, 0.78-1.05). Within racial and ethnic groups, significant associations were observed with different individual vegetables and in women with certain genotypes of isothiocyanate-metabolizing genes.

Conclusions: Vegetable types and isothiocyanate-metabolizing gene polymorphisms affect the associations of cruciferous vegetable intake with breast cancer survival.

Impact: Our findings highlight the importance of considering race and ethnicity when evaluating cruciferous vegetable intake in breast cancer survival.

Background: Globally, Caribbean countries are among the most heavily burdened by both human immunodeficiency virus and cancer. Due to their immunocompromised status, people living with human immunodeficiency virus (PLWH) are more susceptible to human papillomavirus (HPV)-related cancers.

Methods: We conducted a preliminary study evaluating HPV vaccination rates targeting PLWH in Trinidad and Tobago using data from local clinics. This study provided descriptive analysis results, including demographic characteristics of enrolled PLWH and HPV vaccination rates.

Results: A total of 5,615 PLWH (age ranged from 18 to 51 years, with 51.4% women and 48.6% men) were enrolled in this evaluation. Of these, 1,178 patients (21.0%) received HPV vaccines: 22.8% were vaccinated with one dose, 25.6% were vaccinated with two doses, and 51.6% were vaccinated with three doses. The highest uptake of 22.3% was in 2021, followed by 20.1% in 2022, but in 2023, it dropped to 15.5%. Between 2018 and 2020, the uptakes were 13.6% for 2020, 12.7% for 2019, and 5.7% for 2018.

Conclusions: Overall, the HPV vaccination rates among PLWH in Trinidad and Tobago are low; only one in five was vaccinated.

Impact: Results suggest that by implementing comprehensive and targeted programs, clinics have the potential to successfully implement HPV vaccinations toward significantly reducing the incidence and mortality of HPV-related cancers and saving lives. There is a trend with the highest vaccination uptake during the peak years of COVID-19 vaccinations, with HPV vaccination rates almost doubling between 2019 and 2021. Hence, our data suggest that the COVID-19 vaccination program may have boosted HPV vaccination.

Background: Early and precise diagnosis is vital to improving patient outcomes and reducing morbidity. In resource-limited settings, cancer diagnosis is often challenging due to shortages of expert pathologists. We assess the effectiveness of general-purpose pathology foundation models (FM) for the diagnosis and annotation of nonmelanoma skin cancer (NMSC) in resource-limited settings.

Methods: We evaluated three pathology FMs (UNI, PRISM, and Prov-GigaPath) using deidentified NMSC histology images from the Bangladesh Vitamin E and Selenium Trial to predict cancer subtype based on zero-shot whole-slide embeddings. In addition, we evaluated tile aggregation methods and machine learning models for prediction. Lastly, we employed few-shot learning of PRISM tile embeddings to perform whole-slide annotation.

Results: We found that the best model used PRISM's aggregated tile embeddings to train a multilayer perceptron model to predict NMSC subtype [mean area under the receiver operating characteristic curve (AUROC) = 0.925, P < 0.001]. Within the other FMs, we found that using attention-based multi-instance learning to aggregate tile embeddings to train a multilayer perceptron model was optimal (UNI: mean AUROC = 0.913, P < 0.001; Prov-GigaPath: mean AUROC = 0.908, P < 0.001). We finally exemplify the utility of few-shot annotation in computation- and expertise-limited settings.

Conclusions: Our study highlights the important role FMs may play in confronting public health challenges and exhibits a real-world potential for machine learning-aided cancer diagnosis.

Impact: Pathology FMs offer a promising pathway to improve early and precise NMSC diagnosis, especially in resource-limited environments. These tools could also facilitate patient stratification and recruitment for prospective clinical trials aimed at improving NMSC management.

Background: In the United States, breast cancer is common in Asian, Hispanic, and non-Hispanic White (NHW) women, many of whom are immigrants. A healthy lifestyle is vital to survival, but it is unknown how lifestyle varies by nativity among survivors.

Methods: The study included 4,754 racially diverse, female breast cancer survivors from the Northern California, Exploring Networks in a Cohort of Latina and Asian Emigrants, Lifestyle, and Vital Status (ENCLAVE) study. We generated a healthy lifestyle index (HLI) based on World Cancer Research Fund/American Institute for Cancer Research cancer prevention recommendations. Log-binomial regression models, controlling for sociodemographics, were used to evaluate associations between nativity and the highest tertile of HLI, as well as the optimal category of each HLI component.

Results: Foreign-born (vs. US-born) women were more likely to have a high HLI [prevalence ratio (PR) = 1.44; 95% confidence interval (CI), 1.31-1.59]. In stratified models, we observed stronger associations among Hispanic (PR = 1.76; 95% CI, 1.39-2.24) and Asian (PR = 1.60; 95% CI, 1.32-1.94) versus non-Hispanic White (PR = 1.26; 95% CI, 1.08-1.47) women (P-interaction = 0.02). Foreign-born (vs. US-born) women were more likely to have a waist circumference <31.5 inches (PR = 1.19; 95% CI, 1.03-1.37); be normal weight (PR = 1.23; 95% CI, 1.11-1.37); never smoke (PR = 1.11; 95% CI, 1.06-1.17); and consume no sweets (PR = 1.44; 95% CI, 1.21-1.70), low red meat (PR = 1.46; 95% CI, 1.33-1.60), and high fruits and vegetables (PR = 1.46; 95% CI, 1.32-1.62).

Conclusions: Among women diagnosed with breast cancer, foreign-born women had a healthier lifestyle than US-born women.

Impact: Lifestyle differences by nativity in breast cancer survivors may help clarify prognostic differences by nativity.

Background: Prior studies have demonstrated that the overall density of T cells in colorectal tumors is favorably associated with colorectal cancer survival; however, few studies have considered the potentially distinct roles of heterogeneous T-cell subsets in different tissue regions in relation to colorectal cancer outcomes.

Methods: Including 1,113 colorectal cancer tumors from three observational studies, we conducted in situ T-cell profiling using a customized nine-plex [CD3, CD4, CD8, CD45RA, CD45RO, FOXP3, KRT (keratin), MKI67 (Ki-67), and DAPI] multispectral immunofluorescence assay. Multivariable-adjusted Cox proportional hazards models were used to estimate HRs and 95% confidence intervals for the associations of T-cell subset densities in both epithelial and stromal tissue areas in colorectal cancer with disease-specific survival.

Results: Higher CD3+CD4+ and CD3+CD8+ naïve, memory, and regulatory T-cell densities were significantly associated with better colorectal cancer-specific survival in both epithelial and stromal tissue areas (HR highest quantile vs. lowest quantile ranging 0.41-0.68). These associations persisted in models further adjusted for stage at diagnosis and were largely consistent when stratified by microsatellite instability status. However, the further stratification into CD4+ or CD8+ T-cell subsets beyond CD3+ subsets did not significantly improve the performance of our model in explaining colorectal cancer prognosis.

Conclusions: The density of T cells in colorectal cancer tissue, both overall and for several T-cell subset populations, is significantly associated with colorectal cancer-specific survival independent of microsatellite instability status and stage at diagnosis.

Impact: Higher levels of T-cell densities in different locations with different functions are associated with better colorectal cancer-specific survival.