Prognostic models for breast cancer: a systematic review.

BACKGROUND: Breast cancer is the most common cancer in women worldwide, with a great diversity in outcomes among individual patients. The ability to accurately predict a breast cancer outcome is important to patients, physicians, researchers, and policy makers. Many models have been developed and tested in different settings. We systematically reviewed the prognostic models developed and/or validated for patients with breast cancer. METHODS: We conducted a systematic search in four electronic databases and some oncology websites, and a manual search in the bibliographies of the included studies. We identified original studies that were published prior to 1st January 2017, and presented the development and/or validation of models based mainly on clinico-pathological factors to predict mortality and/or recurrence in female breast cancer patients. RESULTS: From the 96 articles selected from 4095 citations found, we identified 58 models, which predicted mortality (n = 28), recurrence (n = 23), or both (n = 7). The most frequently used predictors were nodal status (n = 49), tumour size (n = 42), tumour grade (n = 29), age at diagnosis (n = 24), and oestrogen receptor status (n = 21). Models were developed in Europe (n = 25), Asia (n = 13), North America (n = 12), and Australia (n = 1) between 1982 and 2016. Models were validated in the development cohorts (n = 43) and/or independent populations (n = 17), by comparing the predicted outcomes with the observed outcomes (n = 55) and/or with the outcomes estimated by other models (n = 32), or the outcomes estimated by individual prognostic factors (n = 8). The most commonly used methods were: Cox proportional hazards regression for model development (n = 32); the absolute differences between the predicted and observed outcomes (n = 30) for calibration; and C-index/AUC (n = 44) for discrimination. Overall, the models performed well in the development cohorts but less accurately in some independent populations, particularly in patients with high risk and young and elderly patients. An exception is the Nottingham Prognostic Index, which retains its predicting ability in most independent populations. CONCLUSIONS: Many prognostic models have been developed for breast cancer, but only a few have been validated widely in different settings. Importantly, their performance was suboptimal in independent populations, particularly in patients with high risk and in young and elderly patients.

Personalized treatment of women with early breast cancer: a risk-group specific cost-effectiveness analysis of adjuvant chemotherapy accounting for companion prognostic tests OncotypeDX and Adjuvant!Online.

BACKGROUND: Due to high survival rates and the relatively small benefit of adjuvant therapy, the application of personalized medicine (PM) through risk stratification is particularly beneficial in early breast cancer (BC) to avoid unnecessary harms from treatment. The new 21-gene assay (OncotypeDX, ODX) is a promising prognostic score for risk stratification that can be applied in conjunction with Adjuvant!Online (AO) to guide personalized chemotherapy decisions for early BC patients. Our goal was to evaluate risk-group specific cost effectiveness of adjuvant chemotherapy for women with early stage BC in Austria based on AO and ODX risk stratification. METHODS: A previously validated discrete event simulation model was applied to a hypothetical cohort of 50-year-old women over a lifetime horizon. We simulated twelve risk groups derived from the joint application of ODX and AO and included respective additional costs. The primary outcomes of interest were life-years gained, quality-adjusted life-years (QALYs), costs and incremental cost-effectiveness (ICER). The robustness of results and decisions derived were tested in sensitivity analyses. A cross-country comparison of results was performed. RESULTS: Chemotherapy is dominated (i.e., less effective and more costly) for patients with 1) low ODX risk independent of AO classification; and 2) low AO risk and intermediate ODX risk. For patients with an intermediate or high AO risk and an intermediate or high ODX risk, the ICER is below 15,000 EUR/QALY (potentially cost effective depending on the willingness-to-pay). Applying the AO risk classification alone would miss risk groups where chemotherapy is dominated and thus should not be considered. These results are sensitive to changes in the probabilities of distant recurrence but not to changes in the costs of chemotherapy or the ODX test. CONCLUSIONS: Based on our modeling study, chemotherapy is effective and cost effective for Austrian patients with an intermediate or high AO risk and an intermediate or high ODX risk. In other words, low ODX risk suggests chemotherapy should not be considered but low AO risk may benefit from chemotherapy if ODX risk is high. Our analysis suggests that risk-group specific cost-effectiveness analysis, which includes companion prognostic tests are essential in PM.

Prognostic models in male breast cancer.

PURPOSE: Breast cancer in men is uncommon; it accounts for 1 % of all patients with primary breast cancer. Its treatment is mostly extrapolated from its female counterpart. Accurate predictions are essential for adjuvant systemic treatment decision-making and informing patients. Several predictive models are available for female breast cancer (FBC) including the Morphometric Prognostic Index (MPI), Nottingham Prognostic Index (NPI), Adjuvant! Online and Predict. The aim of this study was to examine and compare the prognostic performance of these models for male breast cancer (MBC). METHODS: The population of this study consists of 166 MBC patients. The prognostic scores of the patients are categorized by good, (moderate) and poor, defined by the test itself (MPI and NPI) or based on tertiles (Adjuvant! Online and Predict). Survival according to prognostic score was compared by Kaplan-Meier analysis and differences were tested by logRank. The prognostic performances were evaluated with C-statistics. Calibration was done with the aim to estimate to what extent the survival rates predicted by Predict were similar to the observed survival rates. RESULTS: All prediction models were capable of discriminating between good, moderate and poor survivors. P-values were highly significant. Comparison between the models using C-statistics (n = 88) showed equal performance of MPI (0.67), NPI (0.68), Adjuvant! Online (0.69) and Predict (0.69). Calibration of Predict showed overestimation for MBC patients. CONCLUSION: In conclusion, MPI, NPI, Adjuvant! and Predict prognostic models, originally developed and validated for FBC patients, also perform quite well for MBC patients.

Transferring MINDACT to Daily Routine: Implementation of the 70-Gene Signature in Luminal Early Breast Cancer - Results from a Prospective Registry of the Austrian Group Medical Tumor Therapy (AGMT).

Background: For hormone receptor (HR)-positive/human epidermal growth factor receptor 2 (HER2)-negative early breast cancer (EBC), adjuvant chemotherapy (ACT) is recommended in the case of high-risk features only. The MINDACT trial showed that patients with high clinical risk (CR) but low genomic risk (GR) defined by the 70-gene signature (MammaPrint®; 70-GS) did not benefit from ACT. In this registry, we investigated the frequency and feasibility of 70-GS and concurrent 80-gene subtyping (BluePrint®) use in HR-positive, HER2-negative EBC. Furthermore, we recorded the frequency of ACT recommendation and the adherence to it when the "MINDACT strategy" was used. Methods: This prospective registry included patients from 2 Austrian cancer centers. Similar to MINDACT, a modified version of Adjuvant!Online was used to determine CR, and 70-GC was used to determine GR in high-CR patients. ACT was recommended to patients with high CR and high GR. Results: Of 224 enrolled patients, 76 (33.9%) had high CR and 67 (88.2%) received genomic testing. Of those, 43 (64.2%) were classified as low and 24 (35.8%) as high GR, respectively. All 24 patients with high CR and GR (10.7% of all patients) received the recommendation for ACT, but ACT was started in only 15 patients (62.5%). The median time from surgery to the start of ACT was 45 days (range 32-68), and the median time from test decision to the test result was 15 days (range 9-56). Conclusion: We showed that the results of the MINDACT trial are reproducible in an Austrian population. Incorporating 70-GS into the daily clinical routine is feasible and mostly accepted by physicians for the guidance of treatment recommendations.

Comparing Genetic Risk and Clinical Risk Classification in Luminal-like Breast Cancer Patients Using a 23-Gene Classifier.

Background: A 23-gene classifier has been developed based on gene expression profiles of Taiwanese luminal-like breast cancer. We aim to stratify risk of relapse and identify patients who may benefit from adjuvant chemotherapy based on genetic model among distinct clinical risk groups. Methods: There were 248 luminal (hormone receptor-positive and human epidermal growth factor receptor II-negative) breast cancer patients with 23-gene classifier results. Using the modified Adjuvant! Online definition, clinical high/low-risk groups were tabulated with the genetic model. The primary endpoint was a recurrence-free interval (RFI) at 5 years. Results: There was a significant difference between the high/low-risk groups defined by the 23-gene classifier for the 5-year prognosis of recurrence (16 recurrences in high-risk and 3 recurrences in low-risk; log-rank test: p < 0.0001). Among the clinically high-risk group, the 5-year RFI of high risk defined by the 23-gene classifier was significantly higher than that of the low-risk group (15 recurrences in high-risk and 2 recurrences in low-risk; log-rank test: p < 0.0001). Conclusion: This study showed that 23-gene classifier can be used to stratify clinically high-risk patients into distinct survival patterns based on genomic risks and displays the potentiality to guide adjuvant chemotherapy. The 23-gene classifier can provide a better estimation of breast cancer prognosis which can help physicians make a better treatment decision.

Adjuvant Chemotherapy for Low-Clinical-Risk Breast Cancer Defined by Modified Version of Adjuvant! Online: A Propensity Score Matched SEER Analysis.

BACKGROUND: The purpose of this research was to investigate whether the modified version of Adjuvant! Online was able to omit chemotherapy (CT) for patients with hormone receptor (HR)-positive, human epidermal growth factor receptor 2 (HER2)-negative, and axillary node-negative breast cancer, who are defined as low clinical risk. METHODS: From 2010 to 2014, HR-positive, HER2-negative, and node-negative breast cancer patients aged 50 years and older were retrieved from the Surveillance, Epidemiology, and End Results (SEER) 18 database. The propensity score matching method was applied between the no-CT and CT groups. Overall survival (OS) was evaluated using Kaplan-Meier analysis and compared across groups using a log-rank test. RESULTS: A total of 48,857 patients were enrolled. After propensity score matching, the numbers of patients in the no-CT and CT groups were both 3,102. The median follow-up period was 37 months. The 5-year OS rates in the no-CT and CT groups were 92 and 91%, respectively (p = 0.066). In the subgroup with a tumor score (tumor size added to tumor grade) of 2-3, OS was significantly higher in the no-CT group than in the CT group (93 vs. 90%, p < 0.001). In the subgroup with a tumor score of 4, OS was not different between these 2 groups (92 vs. 93%, p = 0.47). CONCLUSION: This retrospective study provides evidence that CT may not be beneficial to patients 50 years of age or older with HR-positive, HER2-negative, axillary node-negative breast cancer and additionally defined as low clinical risk by a modified version of Adjuvant! Online.

Decision-making of Adjuvant Chemotherapy for Breast Cancer Patients with Discordant Risk Classifications between Clinical-Pathological Factors and 21-gene Recurrence Score.

Background: Clinical-pathological factors and 21-gene recurrence score (RS) influence adjuvant chemotherapy (ACT) decision for early breast cancer patients. We investigated the decision-making of ACT in patients with discordant risk classifications of clinical-pathological factors and RS. Methods: Patients with hormonal receptor (HR)+/ human epidermal growth factor receptor 2 (HER2)-, early breast cancer, who underwent 21-gene RS testing were identified from Ruijin Hospital (RJBC) and the Surveillance, Epidemiology, and End Results (SEER) database. According to Adjuvant! Online and RS (≤25 or >25), discordant risk classifications were defined as: clinical low-risk/ RS high-risk (C-low/ RS-high) and clinical high-risk/ RS low-risk (C-high/RS-low). McNemar's test was used to assess the changes between pre- and post-RS recommendations. Breast cancer-specific survival (BCSS) was estimated using the Kaplan-Meier methods. Results: Among 727 RJBC patients, the C-low/RS-high group and the C-high/RS-low group represented 19.7% and 21.3% of the cohort. After receiving 21-gene RS results, treatment recommendations were changed for 22.1% patients with discordant risk classifications: ACT rate increased from 41.9% to 75.5% in the C-low/RS-high group and decreased from 63.9% to 60.0% in the C-high/RS-low group. Among 2958 patients from the SEER cohort, 18.4% of the C-high/RS-low group and 59.2% of the C-low/RS-high group received ACT. There was no significant difference in the estimated 3-year BCSS between ACT or not among the C-low/RS-high group (p=0.708) and the C-high/RS-low groups (p=0.391). Conclusion: For patients with discordant risk classifications, physicians were apt to adopt the 21-gene RS rather than routine clinical-pathological factors to guide ACT selection.

Comparison of Nottingham Prognostic Index and Adjuvant Online prognostic tools in young women with breast cancer: review of a single-institution experience.

OBJECTIVE: Accurately predicting the prognosis of young patients with breast cancer (<40 years) is uncertain since the literature suggests they have a higher mortality and that age is an independent risk factor. In this cohort study we considered two prognostic tools; Nottingham Prognostic Index and Adjuvant Online (Adjuvant!), in a group of young patients, comparing their predicted prognosis with their actual survival. SETTING: North East England PARTICIPANTS: Data was prospectively collected from the breast unit at a Hospital in Grimsby between January 1998 and December 2007. A cohort of 102 young patients with primary breast cancer was identified and actual survival data was recorded. The Nottingham Prognostic Index and Adjuvant! scores were calculated and used to estimate 10-year survival probabilities. Pearson's correlation coefficient was used to demonstrate the association between the Nottingham Prognostic Index and Adjuvant! scores. A constant yearly hazard rate was assumed to generate 10-year cumulative survival curves using the Nottingham Prognostic Index and Adjuvant! predictions. RESULTS: Actual 10-year survival for the 92 patients who underwent potentially curative surgery for invasive cancer was 77.2% (CI 68.6% to 85.8%). There was no significant difference between the actual survival and the Nottingham Prognostic Index and Adjuvant! 10-year estimated survival, which was 77.3% (CI 74.4% to 80.2%) and 82.1% (CI 79.1% to 85.1%), respectively. The Nottingham Prognostic Index and Adjuvant! results demonstrated strong correlation and both predicted cumulative survival curves accurately reflected the actual survival in young patients. CONCLUSIONS: The Nottingham Prognostic Index and Adjuvant! are widely used to predict survival in patients with breast cancer. In this study no statistically significant difference was shown between the predicted prognosis and actual survival of a group of young patients with breast cancer.

Cost effectiveness of personalized treatment in women with early breast cancer: the application of OncotypeDX and Adjuvant! Online to guide adjuvant chemotherapy in Austria.

A Breast Cancer Outcomes model was developed at the ONCOTYROL research center to evaluate personalized test-treatment strategies in Austria. The goal was to evaluate the cost-effectiveness of a new 21-gene assay (ODX) when used in conjunction with the Adjuvant! Online (AO) decision aid to support personalized decisions about use of adjuvant chemotherapy in early-stage breast cancer patients in Austria. We applied a validated discrete-event-simulation model to a hypothetical cohort of 50 years old women over a lifetime horizon. The test-treatment strategies of interest were defined using three-letter acronyms. The first (second, third) letter indicates whether patients with a low (intermediate, high) risk according to AO were tested using ODX (Y yes, N no). The main outcomes were life-years gained, quality-adjusted life-years (QALYs), costs and cost effectiveness. Robustness of the results was tested in sensitivity analyses. Results were compared to a Canadian analysis conducted by the Toronto Health Economics and Technology Assessment Collaborative (THETA). Five of eight strategies were dominated (i.e., more costly and less effective: NNY, NYN, YNN, YNY, YYN). The base-case analysis shows that YYY (ODX provided to all patients) is the most effective strategy and is cost effective with an incremental cost-effectiveness ratio of 15,700 EUR per QALY gained. These results are sensitive to changes in the probabilities of distant recurrence, age and costs of chemotherapy. The results of the base-case analysis were comparable to the THETA results. Based on our analyses, using ODX in addition to AO is effective and cost effective in all women in Austria. The development of future genetic tests may require alternative or additional test-treatment strategies to be evaluated.

Correlation between Nottingham Prognostic Index and Adjuvant! Online prognostic tools in patients with early-stage breast cancer in Mid-Western Ireland.

BACKGROUND: Prognostic tools are widely used in the practice of oncology and have been developed to help stratify patients into specific risk-related grouping. We sought to apply tool such tools used for patients with early-stage breast cancer (EBC) and correlate them to actual outcomes. METHODS: A retrospective analysis was designed to include EBC cases seen at the Mid-Western Regional Hospital from January 1, 2002, to December 31, 2002. Information was derived from the patients' records, and indices were derived from prognostic tools. Information was analyzed using descriptive statistics and chi-square or Fisher exact test. RESULTS: A total of 77 patients were found, with a median age of 52.2 years. A median overall survival (OS) of 84 months was observed. The majority presented with moderately differentiated estrogen receptor positive invasive ductal carcinoma and lymph node involvement (60%). Sixty-four percent of patients underwent mastectomy as opposed to breast conservation. Adjuvant cytotoxic chemotherapy uptake was 61%, which was comparable to the proportion of node positive disease. The Nottingham Prognostic Index and Adjuvant! Online (AO) tools were both correlated with actual survival, with the AO showing better correlation. CONCLUSIONS: This report underscores that these predicting tools were both underestimations consistent with the actual OS and highlights the importance of further work in validating these tools within our own population.