Search

Background Although several clinical breast cancer risk models are used to guide screening and prevention, they have only moderate discrimination. Purpose To compare selected existing mammography artificial intelligence (AI) algorithms and the Breast Cancer Surveillance Consortium (BCSC) risk model for prediction of 5-year risk. Materials and Methods This retrospective case-cohort study included data in women with a negative screening mammographic examination (no visible evidence of cancer) in 2016, who were followed until 2021 at Kaiser Permanente Northern California. Women with prior breast cancer or a highly penetrant gene mutation were excluded. Of the 324 009 eligible women, a random subcohort was selected, regardless of cancer status, to which all additional patients with breast cancer were added. The index screening mammographic examination was used as input for five AI algorithms to generate continuous scores that were compared with the BCSC clinical risk score. Risk estimates for incident breast cancer 0 to 5 years after the initial mammographic examination were calculated using a time-dependent area under the receiver operating characteristic curve (AUC). Results The subcohort included 13 628 patients, of whom 193 had incident cancer. Incident cancers in eligible patients (additional 4391 of 324 009) were also included. For incident cancers at 0 to 5 years, the time-dependent AUC for BCSC was 0.61 (95% CI: 0.60, 0.62). AI algorithms had higher time-dependent AUCs than did BCSC, ranging from 0.63 to 0.67 (Bonferroni-adjusted P < .0016). Time-dependent AUCs for combined BCSC and AI models were slightly higher than AI alone (AI with BCSC time-dependent AUC range, 0.66-0.68; Bonferroni-adjusted P < .0016). Conclusion When using a negative screening examination, AI algorithms performed better than the BCSC risk model for predicting breast cancer risk at 0 to 5 years. Combined AI and BCSC models further improved prediction. © RSNA, 2023 Supplemental material is available for this article.

Mammographic density (MD) phenotypes are strongly associated with breast cancer risk and highly heritable. In this GWAS meta-analysis of 24,192 women, we identify 31 MD loci at P

PurposeRe-examine association of fluoxetine and paroxetine with risk of testicular cancer noted in drug screening, with 4 years more follow-up and expanded study of these and other antidepressant drugs.MethodsIn the Kaiser Permanente Medical Care Program in Northern California, 906 men with testicular cancer diagnosed August 1996-December 2010 were compared with 38,253 matched controls with race/ethnicity recorded regarding receipt of antidepressant drugs at least 2 years before diagnosis or control index date. Analyses emphasized duration of use and histological subgroups.ResultsWith control for race/ethnicity and use of other antidepressant drugs, odds ratios (OR) and 95 % confidence intervals (CI) for associations with testicular cancer were as follows: fluoxetine 1.22 (0.88-1.71), paroxetine 1.19 (0.78-1.83), and 1.21 (0.92-1.58) for all serotonin reuptake inhibitors. There was no statistically significant association with risk of all testicular cancers or their histological subtypes for any individual drug or for tricyclics or all antidepressants combined except for citalopram with all testicular cancers 2.55 (1.43-4.52) and those of mixed histology 4.36 (1.50-12.68) and nefazodone with embryonal cancers 9.79 (1.85-51.81). These could readily be chance findings in the context of the many analyses that were performed. Duration of use was not associated with risk of the drugs and drug groups with sufficient numbers of exposed cases for analysis.ConclusionsWe found little evidence to support a testicular carcinogenic effect of fluoxetine, paroxetine, or other antidepressant drugs, but a weakly positive association is not ruled out. The signals in prior screening may have been due to chance and/or uncontrolled confounding.

Background: Hydrochlorothiazide (HCTZ), a common diuretic known to be photosensitizing and previously associated with non‐melanoma skin cancer, was recently reported to be associated with two melanoma subtypes, nodular and lentigo, among residents of Denmark. Our goal was to examine whether Danish findings could be replicated in a US cohort, using a similar study design and analysis. Methods: Among non‐Hispanic White enrollees of Kaiser Permanente Northern California, we conducted an analysis of 9176 melanoma cases and 264 781 controls, matched on age, sex and time in health plan. We examined use of HCTZ prior to cancer diagnosis (cases) or comparable date for controls, categorized as never use, ever use and high use (≥50 000 mg). Electronic health records provided data on prescriptions, cancer diagnoses, and covariates. Conditional logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs), adjusted for education, income and number of dermatology, internal medicine and urgent care visits. Results: We observed a small increase in risk of melanoma, all types combined, associated with high use (≥50 000 mg) of HCTZ (OR = 1.11, 95% CI 1.00–1.23) and no evidence of a dose–response. Risk was more elevated for lentigo subtype (OR = 1.57, 95% CI 1.01–2.42). The somewhat elevated risk for nodular subtype was not statistically significant (OR = 1.22, 95% CI 0.78–1.90). There was very little association of high use with the superficial spreading subtype (OR = 1.05, 95% CI 0.80–1.37). Conclusions: Our findings support a recent report of an association between high use of HCTZ and increased risk of the lentigo subtype of melanoma. [ABSTRACT FROM AUTHOR]