Cancer Models and Real-world Data: Better Together

Data gaps and opportunities for modeling cancer health equity

Population models of cancer reflect the overall US population by drawing on numerous existing data resources for parameter inputs and calibration targets. Models require data inputs that are appropriately representative, collected in a harmonized manner, have minimal missing or inaccurate values, and reflect adequate sample sizes. Data resource priorities for population modeling to support cancer health equity include increasing the availability of data that 1) arise from uninsured and underinsured individuals and those traditionally not included in health-care delivery studies, 2) reflect relevant exposures for groups historically and intentionally excluded across the full cancer control continuum, 3) disaggregate categories (race, ethnicity, socioeconomic status, gender, sexual orientation, etc.) and their intersections that conceal important variation in health outcomes, 4) identify specific populations of interest in clinical databases whose health outcomes have been understudied, 5) enhance health records through expanded data elements and linkage with other data types (eg, patient surveys, provider and/or facility level information, neighborhood data), 6) decrease missing and misclassified data from historically underrecognized populations, and 7) capture potential measures or effects of systemic racism and corresponding intervenable targets for change.

The emerging role of real-world data in advanced breast cancer therapy: Recommendations for collaborative decision-making

Among stakeholders and decision-makers in advanced breast cancer, the demand for insights from real-world data (RWD) is increasing. Although RWD can be used to support decisions throughout different stages of a breast cancer drug's life cycle, barriers exist to its use and acceptance. We propose a collaborative approach to generating and using RWD that is meaningful to multiple stakeholders, and encourage frameworks toward international guidelines to help standardize RWD methodologies to achieve more efficient use of RWD insights.

Cost-effectiveness of treatments for HER2-positive metastatic breast cancer and associated metastases: an overview of systematic reviews

Introduction: Treatment of human epithelial growth factor receptor 2 (HER2)-positive breast cancer has rapidly evolved over the past decades with the addition of trastuzumab, lapatinib, pertuzumab, and trastuzumab emtansine (T-DM1). These treatments have dramatically impacted the survival of HER2-positive metastatic breast cancer (mBC) patients. Nonetheless, these agents are associated with high price tags, begging the question, 'Are treatments for HER2-positive metastatic breast cancer and associated metastases cost-effective'?Areas covered: We examine evidence on the cost-effectiveness of treatments for HER2-positive metastatic breast cancer and associated metastases through a review of systematic reviews on the topic. Additionally, we discuss the implications of our findings and provide recommendations for future directions in the assessment of the cost-effectiveness of targeted directed agents for HER2-positive mBC.Expert opinion: Heterogeneous evidence from cost-effectiveness studies on the use of targeted directed agents for HER2-positive mBC across the world caution against cross-country comparisons of the value of such treatments. It also militates in favor of the production and use of cost-effectiveness analyses for local rather than global decision-making, thus ensuring that economic evaluations reflect the needs of local decision-makers and populations for which they are devised.

Multidimensional Machine Learning Personalized Prognostic Model in an Early Invasive Breast Cancer Population-Based Cohort in China: Algorithm Validation Study

Background

Current online prognostic prediction models for breast cancer, such as Adjuvant! Online and PREDICT, are based on specific populations. They have been well validated and widely used in the United States and Western Europe; however, several validation attempts in non-European countries have revealed suboptimal predictions.

Objective

We aimed to develop an advanced breast cancer prognosis model for disease progression, cancer-specific mortality, and all-cause mortality by integrating tumor, demographic, and treatment characteristics from a large breast cancer cohort in China.

Methods

This study was approved by the Clinical Test and Biomedical Ethics Committee of West China Hospital, Sichuan University on May 17, 2012. Data collection for this project was started in May 2017 and ended in March 2019. Data on 5293 women diagnosed with stage I to III invasive breast cancer between 2000 and 2013 were collected. Disease progression, cancer-specific mortality, all-cause mortality, and the likelihood of disease progression or death within a 5-year period were predicted. Extreme gradient boosting was used to develop the prediction model. Model performance was assessed by calculating the area under the receiver operating characteristic curve (AUROC), and the model was calibrated and compared with PREDICT.

Results

The training, test, and validation sets comprised 3276 (499 progressions, 202 breast cancer-specific deaths, and 261 all-cause deaths within 5-year follow-up), 1405 (211 progressions, 94 breast cancer-specific deaths, and 129 all-cause deaths), and 612 (109 progressions, 33 breast cancer-specific deaths, and 37 all-cause deaths) women, respectively. The AUROC values for disease progression, cancer-specific mortality, and all-cause mortality were 0.76, 0.88, and 0.82 for training set; 0.79, 0.80, and 0.83 for the test set; and 0.79, 0.84, and 0.88 for the validation set, respectively. Calibration analysis demonstrated good agreement between predicted and observed events within 5 years. Comparable AUROC and calibration results were confirmed in different age, residence status, and receptor status subgroups. Compared with PREDICT, our model showed similar AUROC and improved calibration values.

Conclusions

Our prognostic model exhibits high discrimination and good calibration. It may facilitate prognosis prediction and clinical decision making for patients with breast cancer in China.

Delayed Colonoscopy Following a Positive Fecal Test Result and Cancer Mortality

Background

A fecal test followed by diagnostic colonoscopy for a positive result is a widely endorsed screening strategy for colorectal cancer (CRC). However, the relationship between the time delay from the positive test to the follow-up colonoscopy and CRC mortality has not been established.

Methods

From a population-based screening program, we identified CRC patients newly diagnosed from 2005 through 2015 by a positive fecal occult test followed by a colonoscopy. The primary outcome measure was CRC-specific mortality according to four categories for the time elapsed between the positive result and the subsequent colonoscopy.

Results

The 1749 patients underwent colonoscopies within 0–3 months (n = 981, 56.1%), 4–6 months (n = 307, 17.5%), 7–12 months (n = 157, 9.0%), and later than 12 months (n = 304, 17.4%). CRC-specific deaths according to exposure groups were: 13.8% (135 of 981) for 0–3 months, 10.7% (33 of 307) for 4–6 months (crude hazards ratio [HR] = 0.74, 95% confidence interval [CI] = 0.51 to 1.14), 11.5% (18 of 157) for 7–12 months (crude HR = 0.83, 95% CI = 0.51 to 1.42), and 22.7% (69 of 304) for longer than 12 months (crude HR = 1.40, 95% CI = 1.04 to 1.90). The only variable that was associated with mortality risk was the number of positive slides (P = .003). High positivity was twice the value in the 0–3 as the longer-than-12 months group: 51.9% vs 25.0% and similar for the 4–6 and 7–12 months groups (38.1% and 36.5%), respectively. The adjusted HRs for CRC mortality were 0.81 (95% CI = 0.55 to 1.19); 0.83 (95% CI = 0.50 to 1.41), and 1.53 (95% CI = 1.13 to 2.12, P = .006) for the 4–12, 7–12, and longer-than-12-months groups, respectively, compared with the shortest delay group.

Conclusions

Among screen-diagnosed CRC patients, performance of colonoscopy more than 12 months after the initial positive fecal occult blood test was associated with more advanced disease and higher mortality due to CRC.

Validation of a prognostic score for hidden cancer in unprovoked venous thromboembolism

The usefulness of a diagnostic workup for occult cancer in patients with venous thromboembolism (VTE) is controversial. We used the RIETE (Registro Informatizado Enfermedad Trombo Embólica) database to perform a nested case-control study to validate a prognostic score that identifies patients with unprovoked VTE at increased risk for cancer. We dichotomized patients as having low- (≤2 points) or high (≥3 points) risk for cancer, and tried to validate the score at 12 and 24 months. From January 2014 to October 2016, 11,695 VTE patients were recruited. Of these, 1,360 with unprovoked VTE (11.6%) were eligible for the study. At 12 months, 52 patients (3.8%; 95%CI: 2.9–5%) were diagnosed with cancer. Among 905 patients (67%) scoring ≤2 points, 22 (2.4%) had cancer. Among 455 scoring ≥3 points, 30 (6.6%) had cancer (hazard ratio 2.8; 95%CI 1.6–5; p<0.01). C-statistic was 0.63 (95%CI 0.55–0.71). At 24 months, 58 patients (4.3%; 95%CI: 3.3–5.5%) were diagnosed with cancer. Among 905 patients scoring ≤2 points, 26 (2.9%) had cancer. Among 455 patients scoring ≥3 points, 32 (7%) had cancer (hazard ratio 2.6; 95%CI 1.5–4.3; p<0.01). C-statistic was 0.61 (95%CI, 0.54–0.69). We validated our prognostic score at 12 and 24 months, although prospective cohort validation is needed. This may help to identify patients for whom more extensive screening workup may be required.

What cervical screening is appropriate for women who have been vaccinated against high risk HPV? A simulation study

Women vaccinated against HPV16/18 are approaching the age for cervical screening; however, an updated screening algorithm has not been agreed. We use a microsimulation model calibrated to real published data to determine the appropriate screening intensity for vaccinated women. Natural histories in the absence of vaccination were simulated for 300,000 women using 10,000 sets of transition probabilities. Vaccination with (i) 100% efficacy against HPV16/18, (ii) 15% cross-protection, (iii) 22% cross-protection, (iv) waning vaccine efficacy and (v) 100% efficacy against HPV16/18/31/33/45/52/58 was added, as were a range of screening scenarios appropriate to the UK. To benchmark cost-benefits of screening for vaccinated women, we evaluated the proportion of cancers prevented per additional screen (incremental benefit) of current cytology and likely HPV screening scenarios in unvaccinated women. Slightly more cancers are prevented through vaccination with no screening (70.3%, 95% CR: 65.1-75.5) than realistic compliance to the current UK screening programme in the absence of vaccination (64.3%, 95% CR: 61.3-66.8). In unvaccinated women, when switching to HPV primary testing, there is no loss in effectiveness when doubling the screening interval. Benchmarking supports screening scenarios with incremental benefits of ≥2.0%, and rejects scenarios with incremental benefits ≤0.9%. In HPV16/18-vaccinated women, the incremental benefit of offering a third lifetime screen was at most 3.3% (95% CR: 2.2-4.5), with an incremental benefit of 1.3% (-0.3-2.8) for a fourth screen. For HPV16/18/31/33/45/52/58-vaccinated women, two lifetime screens are supported. It is important to know women's vaccination status; in these simulations, HPV16/18-vaccinated women require three lifetime screens, HPV16/18/31/33/45/52/58-vaccinated women require two lifetime screens, yet unvaccinated women require seven lifetime screens.

Cost-Effectiveness of Breast Cancer Screening in Turkey, a Developing Country: Results from Bahçeşehir Mammography Screening Project

OBJECTIVE

We used the results from the first three screening rounds of Bahcesehir Mammography Screening Project (BMSP), a 10-year (2009-2019) and the first organized population-based screening program implemented in a county of Istanbul, Turkey, to assess the potential cost-effectiveness of a population-based mammography screening program in Turkey.

MATERIALS AND METHODS

Two screening strategies were compared: BMSP (includes three biennial screens for women between 40-69) and Turkish National Breast Cancer Registry Program (TNBCRP) which includes no organized population-based screening. Costs were estimated using direct data from the BMSP project and the reimbursement rates of Turkish Social Security Administration. The life-years saved by BMSP were estimated using the stage distribution observed with BMSP and TNBCRP.

RESULTS

A total of 67 women (out of 7234 screened women) were diagnosed with breast cancer in BMSP. The stage distribution for AJCC stages O, I, II, III, IV was 19.4%, 50.8%, 20.9%, 7.5%, 1.5% and 4.9%, 26.6%, 44.9%, 20.8%, 2.8% with BMSP and TNBCRP, respectively. The BMSP program is expected to save 279.46 life years over TNBCRP with an additional cost of $677.171, which implies an incremental cost-effectiveness ratio (ICER) of $2.423 per saved life year. Since the ICER is smaller than the Gross Demostic Product (GDP) per capita in Turkey ($10.515 in 2014), BMSP program is highly cost-effective and remains cost-effective in the sensitivity analysis.

CONCLUSION

Mammography screening may change the stage distribution of breast cancer in Turkey. Furthermore, an organized population-based screening program may be cost-effective in Turkey and in other developing countries. More research is needed to better estimate life-years saved with screening and further validate the findings of our study.

Cancer Models and Real-world Data: Better Together

Decision-analytic models are increasingly used to inform health policy decisions. These models synthesize available data on disease burden and intervention effectiveness to project estimates of the long-term consequences of care, which are often absent when clinical or policy decisions must be made. While models have been influential in informing US cancer screening guidelines under ideal conditions, incorporating detailed data on real-world screening practice has been limited given the complexity of screening processes and behaviors throughout diverse health delivery systems in the United States. We describe the synergies that exist between decision-analytic models and health care utilization data that are increasingly accessible through research networks that assemble data from the growing number of electronic medical record systems. In particular, we present opportunities to enrich cancer screening models by grounding analyses in real-world data with the goals of projecting the harms and benefits of current screening practices, evaluating the value of existing and new technologies, and identifying the weakest links in the cancer screening process where efforts for improvement may be most productively focused. We highlight the example of the National Cancer Institute-funded consortium Population-based Research Optimizing Screening through Personalized Regimens (PROSPR), a collaboration to harmonize and analyze screening process and outcomes data on breast, colorectal, and cervical cancers across seven research centers. The pairing of models with such data can create more robust models to not only better inform policy but also inform health care systems about best approaches to improve the provision of cancer screening in the United States.

Targeted Cancer Screening in Average-Risk Individuals

Targeted cancer screening refers to use of disease risk information to identify those most likely to benefit from screening. Researchers have begun to explore the possibility of refining screening regimens for average-risk individuals using genetic and non-genetic risk factors and previous screening experience. Average-risk individuals are those not known to be at substantially elevated risk, including those without known inherited predisposition, without comorbidities known to increase cancer risk, and without previous diagnosis of cancer or pre-cancer. In this paper, we describe the goals of targeted cancer screening in average-risk individuals, present factors on which cancer screening has been targeted, discuss inclusion of targeting in screening guidelines issued by major U.S. professional organizations, and present evidence to support or question such inclusion. Screening guidelines for average-risk individuals currently target age; smoking (lung cancer only); and, in some instances, race; family history of cancer; and previous negative screening history (cervical cancer only). No guidelines include common genomic polymorphisms. RCTs suggest that targeting certain ages and smoking histories reduces disease-specific cancer mortality, although some guidelines extend ages and smoking histories based on statistical modeling. Guidelines that are based on modestly elevated disease risk typically have either no or little evidence of an ability to affect a mortality benefit. In time, targeted cancer screening is likely to include genetic factors and past screening experience as well as non-genetic factors other than age, smoking, and race, but it is of utmost importance that clinical implementation be evidence-based.

Registry-based study of trends in breast cancer screening mammography before and after the 2009 U.S. Preventive Services Task Force recommendations

PURPOSE

To determine whether the 2009 U.S. Preventive Services Task Force (USPSTF) guidelines for breast cancer mammography screening were followed by changes in screening utilization in the state of Vermont.

MATERIALS AND METHODS

This retrospective study was HIPAA compliant and approved by the institutional review board, with waiver of informed consent. Trends in screening mammography utilization during 1997-2011 were examined among approximately 150,000 women aged 40 years and older in the state of Vermont using statewide mammography registry data.

RESULTS

The percentage of Vermont women aged 40 years and older screened in the past year declined from 45.3% in 2009% to 41.6% in 2011 (an absolute decrease of -3.7 percentage points; 95% confidence interval [CI]: -3.3, -4.1). The largest decline in utilization was among women aged 40-49 years (-4.8 percentage points; 95% CI: -4.1, -5.4), although substantial declines were also observed among women aged 50-74 years (-3.0 percentage points; 95% CI: -2.6, -3.5) and women aged 75 years and older (-3.1 percentage points; 95% CI: -2.3, -4.0). The percentage of women aged 50-74 years screened within the past 2 years declined by -3.4 percentage points (95% CI: -3.0, -3.9) from 65.4% in 2009 to 61.9% in 2011.

CONCLUSION

After years of increasing screening mammography utilization in Vermont, there was a decline in screening, which coincided with the release of the 2009 USPSTF recommendations. The age-specific patterns in utilization were generally consistent with the USPSTF recommendations, although there was also evidence that the percentage of women aged 50-74 years screened in the past 2 years declined since 2009.