Individually tailored screening of breast cancer with genes, tumour phenotypes, clinical attributes, and conventional risk factors

Multistate models for the natural history of cancer progression

BACKGROUND

Multistate models can be effectively used to characterise the natural history of cancer. Inference from such models has previously been useful for setting screening policies.

METHODS

We introduce the basic elements of multistate models and the challenges of applying these models to cancer data. Through simulation studies, we examine (1) the impact of assuming time-homogeneous Markov transition intensities when the intensities depend on the time since entry to the current state (i.e., the process is time-inhomogenous semi-Markov) and (2) the effect on precancer risk estimation when observation times depend on an unmodelled intermediate disease state.

RESULTS

In the settings we examined, we found that misspecifying a time-inhomogenous semi-Markov process as a time-homogeneous Markov process resulted in biased estimates of the mean sojourn times. When screen-detection of the intermediate disease leads to more frequent future screening assessments, there was minimal bias induced compared to when screen-detection of the intermediate disease leads to less frequent screening.

CONCLUSIONS

Multistate models are useful for estimating parameters governing the process dynamics in cancer such as transition rates, sojourn time distributions, and absolute and relative risks. As with most statistical models, to avoid incorrect inference, care should be given to use the appropriate specifications and assumptions.

Precision Science on Incidence and Progression of Early-Detected Small Breast Invasive Cancers by Mammographic Features

The aim was to evaluate how the inter-screening interval affected the performance of screening by mammographic appearances. This was a Swedish retrospective screening cohort study with information on screening history and mammography features in two periods (1977-1985 and 1996-2010). The pre-clinical incidence and the mean sojourn time (MST) for small breast cancer allowing for sensitivity by mammographic appearances were estimated. The percentage of interval cancer against background incidence (I/E ratio) was used to assess the performance of mammography screening by different inter-screening intervals. The sensitivity-adjusted MSTs (in years) were heterogeneous with mammographic features, being longer for powdery and crushed stone-like calcifications (4.26, (95% CI, 3.50-5.26)) and stellate masses (3.76, (95% CI, 3.15-4.53)) but shorter for circular masses (2.65, (95% CI, 2.06-3.55)) in 1996-2010. The similar trends, albeit longer MSTs, were also noted in 1977-1985. The I/E ratios for the stellate type were 23% and 32% for biennial and triennial screening, respectively. The corresponding figures were 32% and 43% for the circular type and 21% and 29% for powdery and crushed stone-like calcifications, respectively. Mammography-featured progressions of small invasive breast cancer provides a new insight into personalized quality assurance, surveillance, treatment and therapy of early-detected breast cancer.

Personalized breast cancer screening strategies: A systematic review and quality assessment

BACKGROUND

The effectiveness of breast cancer screening is still under debate. Our objective was to systematically review studies assessing personalized breast cancer screening strategies based on women's individual risk and to conduct a risk of bias assessment.

METHODS

We followed the standard methods of The Cochrane Collaboration and PRISMA declaration and searched the MEDLINE, EMBASE and Clinical Trials databases for studies published in English. The quality of the studies was assessed using the ISPOR-AMCP-NPC Questionnaire and The Cochrane Risk of Bias Tool. Two independent reviewers screened full texts and evaluated the risk of bias.

RESULTS

Out of the 1533 initially retrieved citations, we included 13 studies. Three studies were randomized controlled trials, while nine were mathematical modeling studies, and one was an observational pilot study. The trials are in the recruitment phase and have not yet reported their results. All three trials used breast density and age to define risk groups, and two of them included family history, previous biopsies, and genetic information. Among the mathematical modeling studies, the main risk factors used to define risk groups were breast density, age, family history, and previous biopsies. Six studies used genetic information to define risk groups. The most common outcome measures were the gain in quality-adjusted life years (QALY), absolute costs, and incremental cost-effectiveness ratio (ICER), while the main outcome in the observational study was the detection rate. In all models, personalized screening strategies were shown to be effective. The randomized trials were of good quality. The modeling studies showed moderate risk of bias but there was wide variability across studies. The observational study showed a low risk of bias but its utility was moderate due to its pilot design and its relatively small scale.

CONCLUSIONS

There is some evidence of the effectiveness of screening personalization in terms of QUALYs and ICER from the modeling studies and the observational study. However, evidence is lacking on feasibility and acceptance by the target population.

REVIEW REGISTRATION

PROSPERO: CRD42018110483.

Bayesian measurement-error-driven hidden Markov regression model for calibrating the effect of covariates on multistate outcomes: Application to androgenetic alopecia

Multistate Markov regression models used for quantifying the effect size of state‐specific covariates pertaining to the dynamics of multistate outcomes have gained popularity. However, the measurements of multistate outcome are prone to the errors of classification, particularly when a population‐based survey/research is involved with proxy measurements of outcome due to cost consideration. Such a misclassification may affect the effect size of relevant covariates such as odds ratio used in the field of epidemiology. We proposed a Bayesian measurement‐error‐driven hidden Markov regression model for calibrating these biased estimates with and without a 2‐stage validation design. A simulation algorithm was developed to assess various scenarios of underestimation and overestimation given nondifferential misclassification (independent of covariates) and differential misclassification (dependent on covariates). We applied our proposed method to the community‐based survey of androgenetic alopecia and found that the effect size of the majority of covariate was inflated after calibration regardless of which type of misclassification. Our proposed Bayesian measurement‐error‐driven hidden Markov regression model is practicable and effective in calibrating the effects of covariates on multistate outcome, but the prior distribution on measurement errors accrued from 2‐stage validation design is strongly recommended.

Assessment of the effects of decision aids about breast cancer screening: a systematic review and meta-analysis

OBJECTIVE

The aim of this systematic review and meta-analysis of randomised controlled trials (RCTs) and observational studies is to assess the effect of decision aids (DAs) in women aged 50 and below facing the decision to be screened for breast cancer.

SETTING

Screening for breast cancer.

INTERVENTION

DAs aimed to help women make a deliberative choice regarding participation in mammography screening by providing information on the options and outcomes.

ELIGIBLE STUDIES

We included published original, non-pilot, studies that assess the effect of DAs for breast cancer screening. We excluded the studies that evaluated only participation intention or actual uptake. The studies' risk of bias was assessed with the Cochrane Collaboration's tool for RCTs and the National Institutes of Health Quality Assessment Tool for non-RCTs.

PRIMARY AND SECONDARY OUTCOMES

The main outcome measures were informed choice, decisional conflict and/or confidence, and knowledge. Secondary outcomes were values, attitudes, uncertainty and intention to be screened.

RESULTS

A total of 607 studies were identified, but only 3 RCTs and 1 before-after study were selected. The use of DAs increased the proportion of women making an informed decision by 14%, 95% CI (2% to 27%) and the proportion of women with adequate knowledge by 12%, 95% CI (7% to 16%). We observed heterogeneity among the studies in confidence in the decision. The meta-analysis of the RCTs showed a significant decrease in confidence in the decision and in intention to be screened.

CONCLUSIONS

Tools to aid decision making in screening for breast cancer improve knowledge and promote informed decision; however, we found divergent results on decisional conflict and confidence in the decision. Under the current paradigm change, which favours informed choice rather than maximising uptake, more research is necessary for the improvement of DAs.

Initiators and promoters for the occurrence of screen-detected breast cancer and the progression to clinically-detected interval breast cancer

Background

The risk factors responsible for breast cancer have been well documented, but the roles of risk factors as initiators, causing the occurrence of screen-detected breast cancer, or promoters, responsible for the progression of the screen-detected to the clinically-detected breast cancer, have been scarcely evaluated.

Methods

We used data from women in a cohort in Kopparberg (Dalarna), Sweden between 1977 and 2010. Conventional risk factors, breast density, and tumor-specific biomarkers are superimposed to the temporal course of the natural history of the disease.

Results

The results show that older age at first full-term pregnancy, dense breast, and a family history of breast cancer increased the risk of entering the preclinical screen-detectable phase of breast cancer by 23%, 41%, and 89%, respectively. Overweight/obesity (body mass index ≥25 kg/m²) was a significant initiator (adjusted relative risk [aRR] 1.15; 95% confidence interval [CI], 0.99–1.33), but was inversely associated with the role of promoter (aRR 0.65; 95% CI, 0.51–0.82). Dense breast (aRR 1.46; 95% CI, 1.12–1.91), triple-negative (aRR 2.07; 95% CI, 1.37–3.15), and Ki-67 positivity (aRR 1.66; 95% CI, 1.19–2.30) were statistically significant promoters. When the molecular biomarkers were considered collectively as one classification, the basal-like subtype was the most influential subtype on promoters (aRR 4.24; 95% CI, 2.56–7.02) compared with the Luminal A subtype.

Discussion

We ascertained state-dependent covariates of initiators and promoters to classify the risk of the two-step progression of breast cancer. The results of the current study are useful for individually-tailored screening and personalized clinical surveillance of patients with breast cancer that was detected at an early stage.

•

Identify state-dependent initiators and promoters for breast cancer.

•

Build a multistage risk assessment model for individual risk stratification.

•

Facilitate individually-tailored screening and clinical surveillance.

Prostate cancer screening using risk stratification based on a multi-state model of genetic variants

BACKGROUND

Risk-stratified screening for prostate cancer (PCa) with prostate-specific antigen (PSA) testing incorporating genetic variants has received some attention but has been scarcely investigated. We developed a model to stratify the Finnish population by different risk profiles related to genetic variants to optimize the screening policy.

METHODS

Data from the Finnish randomized controlled trial on screening for PCa with PSA testing were used to estimate a six-state Markov model of disease progression. Blood samples from Finnish men were used to assess the risk of PCa related to three genetic variants (rs4242382, rs138213197, and rs200331695). A risk score-based approach combined with a series of computer simulation models was applied to optimize individual screening policies.

RESULTS

The 10-year risk of having progressive prostate cancer detected ranged from 43% in the top 5% risk group to approximately 11% in the bottom half of the population. Using the median group, with screening every four years beginning at 55 years-old, as the reference group, the recommended age beginning screening was approximately 47 years-old for the top 5% risk group and 55 years-old for those in the lower 60% risk group. The recommended interscreening interval has been shortened for individuals in the high risk group. The increased availability of genomic information allows the proposed multistate model to be more discriminating with respect to risk stratification and the suggested screening policy, particularly for the lowest risk groups-. --

CONCLUSIONS

A multi-state genetic variant-based model was developed for further application to population risk stratification to optimize the interscreening interval and the age at which to begin screening for PSA. A small sub-group of the population is likely to benefit from more intensive screening with early start and short interval, while half of the population is unlikely to benefit from such protocol (compared with four-year interval after age 55 years).

Is mammography screening history a predictor of future breast cancer risk?

Inspired by the model by Walter and Day for risk of cervical cancer following negative screens, one might hypothesize that women in a mammography screening programme with a certain number of negative screens had a lower remaining breast cancer risk than that of women in general. We studied whether number of negative screens was a predictor for a low remaining breast cancer risk in women participating in the mammography screening programmes in Stockholm, Copenhagen and Funen. Data were collected from the mammography screening programmes in Stockholm, Sweden (1989-2012), Copenhagen, Denmark (1991-2009) and Funen, Denmark (1993-2009), and linked to the respective cancer registries. We calculated cumulative hazard rates for breast cancer in women in cohorts defined by age at entry and number of negative screens for the maximum follow-up period in each screening centre. For all centres and cohorts, the cumulative hazard were parallel for all number of negative screens, from after the time, when the women were scheduled to be invited for the next screen. This means that the remaining breast cancer risk is similar no matter how many negative screens a woman have had. Number of negative screens was not a predictor of a low remaining breast cancer risk in women participating in the mammography screening programmes in Stockholm, Sweden, Copenhagen and Funen, Denmark. The history of previous negative screens is therefore not suitable for personalisation of mammography screening.

Metaplastic breast carcinomas - analysis of prognostic factors in a case series

Aim of the study

Metaplastic breast carcinomas (MBC) are a rare group of cancers, accounting for about 1% of all breast cancers. The study presents a case series of MBC patients diagnosed, treated and followed up in one healthcare center.

Material and methods

The study group comprised 18 women at the median age of 63 years. The most common carcinoma type in the study group was MBC with squamous epithelial differentiation (56%). Estrogen receptor expression was identified in one patient. No steroid or HER2 receptor expression was found in the remaining patients. We analyzed recurrence and survival rates in relation to clinical and therapeutic factors by using the Kaplan-Meier method.

Results

A significantly longer overall survival time was noted among patients treated with adjuvant radiation therapy, p = 0.018. No other factors had a significant influence on survival. Because of the small size of the study group, results obtained in the study should be treated with caution.