Real-World Impact of Comprehensive Genomic Profiling on Biomarker Detection, Receipt of Therapy, and Clinical Outcomes in Advanced Non-Small Cell Lung Cancer.

PURPOSE: Therapeutic decision making for patients with advanced non-small cell lung cancer (aNSCLC) includes a growing number of options for genomic, biomarker-guided, targeted therapies. We compared actionable biomarker detection, targeted therapy receipt, and real-world overall survival (rwOS) in patients with aNSCLC tested with comprehensive genomic profiling (CGP) versus small panel testing (SP) in real-world community health systems. METHODS: Patients older than 18 years diagnosed with aNSCLC between January 1, 2015, and December 31, 2020, who received biomarker testing were followed until death or study end (September 30, 2021), and categorized by most comprehensive testing during follow-up: SP (≤52 genes) or CGP (>52 genes). RESULTS: Among 3,884 patients (median age, 68 years; 50% female; 73% non-Hispanic White), 20% received CGP and 80% SP. The proportion of patients with ≥one actionable biomarker (actionability) was significantly higher in CGP than in SP (32% v 14%; P < .001). Of patients with actionability, 43% (CGP) and 38% (SP) received matched therapies (P = .20). Among treated patients, CGP before first-line treatment was associated with higher likelihood of matched therapy in any line (odds ratio, 3.2 [95% CI, 1.84 to 5.53]). CGP testing (hazard ratio [HR], 0.80 [95% CI, 0.72 to 0.89]) and actionability (HR, 0.84 [95% CI, 0.77 to 0.91]) were associated with reduced risk of mortality. Among treated patients with actionability, matched therapy receipt showed improved median rwOS in months in CGP (34 [95% CI, 21 to 49] matched v 14 [95% CI, 10 to 18] unmatched) and SP (27 [95% CI, 21 to 43] matched v 10 [95% CI, 8 to 14] unmatched). CONCLUSION: Patients who received CGP had improved detection of actionable biomarkers and greater use of matched therapies, both of which were associated with significant increases in survival.

System dynamics simulation for evaluating implementation strategies of genomic sequencing: tutorial and conceptual model.

INTRODUCTION: Precision Medicine (PM), especially in oncology, involve diagnostic and complex treatment pathways that are based on genomic features. To conduct evaluation and decision analysis for PM, advanced modeling techniques are needed due to its complexity. Although System Dynamics (SD) has strong modeling power, it has not been widely used in PM and individualized treatment. AREAS COVERED: We explained SD tools using examples in cancer context and the rationale behind using SD for genomic testing and personalized oncology. We compared SD with other Dynamic Simulation Modelling (DSM) methods and listed SD's advantages. We developed a conceptual model using Causal Loop Diagram (CLD) for strategic decision-making in Whole Genome Sequencing (WGS) implementation. EXPERT OPINION: The paper demonstrates that SD is well-suited for health policy evaluation challenges and has useful tools for modeling precision oncology and genomic testing. SD's system-oriented modeling captures dynamic and complex interactions within systems using feedback loops. SD models are simple to implement, utilize less data and computational resources, and conduct both exploratory and explanatory analyses over time. If the targeted system has complex interactions and many components, deals with lack of data, and requires interpretability and clinicians' input, SD offers attractive advantages for modeling and evaluating scenarios.

Integrating a Polygenic Risk Score for Coronary Artery Disease as a Risk-Enhancing Factor in the Pooled Cohort Equation: A Cost-Effectiveness Analysis Study.

Background Cardiovascular diseases are the leading cause of death in the United States, yet a significant proportion of adults at high risk remain undetected by standard screening practices. Polygenic risk score for coronary artery disease (CAD-PRS) improves precision in determining the 10-year risk of atherosclerotic cardiovascular disease but health benefits and health care costs associated with CAD-PRS are unknown. We examined the cost-effectiveness of including CAD-PRS as a risk-enhancing factor in the pooled cohort equation (PCE)-the standard of care for determining the risk of atherosclerotic cardiovascular disease-versus PCE alone. Methods and Results We applied a Markov model on a cohort of 40-year-old individuals with borderline or intermediate 10-year risk (5% to <20%) for atherosclerotic cardiovascular disease to identify those in the top quintile of the CAD-PRS distribution who are at high risk and eligible for statin prevention therapy. Health outcomes examined included coronary artery disease (CAD; ie, myocardial infarction) and ischemic stroke. The model projected medical costs (2019 US$) of screening for CAD, statin prevention therapy, treatment, and monitoring patients living with CAD or ischemic stroke and quality-adjusted life-years for PCE+CAD-PRS versus PCE alone. Deterministic and probabilistic sensitivity analyses and scenario analyses were performed to examine uncertainty in parameter inputs. PCE+CAD-PRS was dominant compared with PCE alone in the 5- and 10-year time horizons. We found that, respectively, PCE+CAD-PRS had 0.003 and 0.011 higher mean quality-adjusted life-years and $40 and $181 lower mean costs per person screened, with 29 and 50 fewer events of CAD and ischemic stroke in a cohort of 10 000 individuals compared with PCE alone. The risk of developing CAD, the effectiveness of statin prevention therapy, and the cost of treating CAD had the largest impact on the cost per quality-adjusted life-year gained. However, this cost remained below the $50 000 willingness-to-pay threshold except when the annual risk of developing CAD was <0.006 in the 5-year time horizon. Results from Monte Carlo simulation indicated that PCE+CAD-PRS would be cost-effective. with the probability of 94% and 99% at $50 000 willingness-to-pay threshold in the 5- and 10-year time horizon, respectively. Conclusions Implementing CAD-PRS as a risk-enhancing factor in the PCE to determine the risk of atherosclerotic cardiovascular disease reduced the mean cost per individual, improved quality-adjusted life-years, and averted future events of CAD and ischemic stroke when compared with PCE alone.

Cost-effectiveness of exome and genome sequencing for children with rare and undiagnosed conditions.

PURPOSE: This study aimed to estimate the cost-effectiveness of exome sequencing (ES) and genome sequencing (GS) for children. METHODS: We modeled costs, diagnoses, and quality-adjusted life years (QALYs) for diagnostic strategies for critically ill infants (aged <1 year) and children (aged <18 years) with suspected genetic conditions: (1) standard of care (SOC) testing, (2) ES, (3) GS, (4) SOC followed by ES, (5) SOC followed by GS, (6) ES followed by GS, and (7) SOC followed by ES followed by GS. We calculated the 10-year incremental cost per additional diagnosis, and lifetime incremental cost per QALY gained, from a health care perspective. RESULTS: First-line GS costs $15,048 per diagnosis vs SOC for infants and $27,349 per diagnosis for children. If GS is unavailable, ES represents the next most efficient option compared with SOC ($15,543 per diagnosis for infants and $28,822 per diagnosis for children). Other strategies provided the same or fewer diagnoses at a higher incremental cost per diagnosis. Lifetime results depend on the patient's assumed long-term prognosis after diagnosis. For infants, GS ranged from cost-saving (vs all alternatives) to $18,877 per QALY (vs SOC). For children, GS (vs SOC) ranged from $119,705 to $490,047 per QALY. CONCLUSION: First-line GS may be the most cost-effective strategy for diagnosing infants with suspected genetic conditions. For all children, GS may be cost-effective under certain assumptions. ES is nearly as efficient as GS and hence is a viable option when GS is unavailable.

Cost-effectiveness of genome sequencing for diagnosing patients with undiagnosed rare genetic diseases.

PURPOSE: To estimate the cost-effectiveness of genome sequencing (GS) for diagnosing critically ill infants and noncritically ill pediatric patients (children) with suspected rare genetic diseases from a United States health sector perspective. METHODS: A decision-analytic model was developed to simulate the diagnostic trajectory of patients. Parameter estimates were derived from a targeted literature review and meta-analysis. The model simulated clinical and economic outcomes associated with 3 diagnostic pathways: (1) standard diagnostic care, (2) GS, and (3) standard diagnostic care followed by GS. RESULTS: For children, costs of GS ($7284) were similar to that of standard care ($7355) and lower than that of standard care followed by GS pathways ($12,030). In critically ill infants, when cost estimates were based on the length of stay in the neonatal intensive care unit, the lowest cost pathway was GS ($209,472). When only diagnostic test costs were included, the cost per diagnosis was $17,940 for standard, $17,019 for GS, and $20,255 for standard care followed by GS. CONCLUSION: The results of this economic model suggest that GS may be cost neutral or possibly cost saving as a first line diagnostic tool for children and critically ill infants.

The diagnostic trajectory of infants and children with clinical features of genetic disease.

We characterized US pediatric patients with clinical indicators of genetic diseases, focusing on the burden of disease, utilization of genetic testing, and cost of care. Curated lists of diagnosis, procedure, and billing codes were used to identify patients with clinical indicators of genetic disease in healthcare claims from Optum's de-identified Clinformatics® Database (13,076,038 unique patients). Distinct cohorts were defined to represent permissive and conservative estimates of the number of patients. Clinical phenotypes suggestive of genetic diseases were observed in up to 9.4% of pediatric patients and up to 44.7% of critically-ill infants. Compared with controls, patients with indicators of genetic diseases had higher utilization of services (e.g., mean NICU length of stay of 31.6d in a cohort defined by multiple congenital anomalies or neurological presentations compared with 10.1d for patients in the control population (P < 0.001)) and higher overall costs. Very few patients received any genetic testing (4.2-8.4% depending on cohort criteria). These results highlight the substantial proportion of the population with clinical features associated with genetic disorders and underutilization of genetic testing in these populations.

Economic Impact of Coverage Expansion for Non-invasive Prenatal Testing Through a Performance-Based Risk-Sharing Agreement.

BACKGROUND: Harvard Pilgrim Health Care expanded coverage for non-invasive prenatal testing (NIPT) to include all pregnant, single-gestation women aged < 35 years, through a performance-based risk-sharing (PBRS) agreement with Illumina to offset costs from coverage expansion. NIPT analyzes cell-free DNA fragments from a maternal blood sample to screen for fetal aneuploidies and is considered a more accurate screening method than conventional serum biochemical screening and nuchal translucency ultrasound-based approaches. OBJECTIVE: This study assessed the impact of NIPT coverage expansion on prenatal screening strategies and payer expenditures. METHODS: This was a real-world comparison of utilization and expenditures of prenatal screening and diagnostic testing in pregnant women aged < 35 years pre- (1 March 2016-28 February 2018) and post- (1 March 2018-30 September 2019) coverage expansion. Incidence rate ratios (IRRs) with 95% confidence intervals (CIs) were estimated to compare changes in utilization of conventional and NIPT-based prenatal screening methods. Change in per member per month (PMPM) expenditures in $US year 2020 were assessed post-coverage expansion using a budget impact model. RESULTS: A total of 5041 and 4109 distinct pregnancies were identified in pre- and post-coverage expansion periods, respectively. Mean ± standard deviation maternal age was consistent between pre- and post-coverage expansion periods (30.35 ± 3.35 and 30.33 ± 3.28, respectively). Screening orders for conventional methods decreased, with an adjusted IRR in the post-expansion period of 0.87 (95% CI 0.85-0.90) times the rate in the pre-expansion period; orders for NIPT increased, with an adjusted IRR in the post-expansion period of 1.41 (95% CI 1.32-1.51) times the rate in the pre-expansion period. Invasive diagnostic testing was low at baseline (1.0%) and did not change post-coverage expansion. The change in PMPM is estimated at $US0.026 post-coverage expansion. CONCLUSION: The PBRS agreement to expand NIPT coverage for women aged < 35 years was associated with an increase in NIPT utilization, decreases in conventional screening methods, and a modest increase in PMPM expenditures.

Multi-Institutional, Prospective Clinical Utility Study Evaluating the Impact of the 92-Gene Assay (CancerTYPE ID) on Final Diagnosis and Treatment Planning in Patients With Metastatic Cancer With an Unknown or Unclear Diagnosis.

PURPOSE: Metastatic cancers of unknown primary or with unclear diagnoses pose diagnostic and management challenges, often leading to poor outcomes. Studies of the 92-gene assay have demonstrated improved diagnostic accuracy compared with standard pathology techniques and improved survival in patients treated on the basis of assay results. The current study assessed the clinical impact of the 92-gene assay on diagnostic and treatment decisions for patients with unknown or uncertain diagnoses. METHODS: Patients in this prospective, multi-institutional, decision-impact study included those for whom the 92-gene assay was ordered as part of routine care. Participating physicians completed electronic case report forms that contained standardized, specialty-specific questionnaires. Data collection included patient and tumor characteristics and clinical history. The key study objective of clinical impact was calculated on the basis of changes in final diagnosis and treatment after testing. RESULTS: Data collection included 444 patients, 107 physicians (73 oncologists and 34 pathologists), and 28 sites. Molecular diagnoses from 22 different tumor types and subtypes across all cases were provided in 95.5% of patients with a reportable result (n = 397). Physicians reported that the 92-gene assay was used broadly for diagnostic dilemmas that ranged from single suspected tumor type (29%) to a differential diagnosis of two or more suspected tumor types (30%) or cancers of unknown primary (41%). Integration of 92-gene assay results led to a change in the recommended treatment in 47% of patients. CONCLUSION: Findings from this clinical utility study demonstrate that the 92-gene assay led to a change in treatment decisions in every other patient case. These data additionally define the role of this assay in clinical practice and strongly support the consideration of molecular tumor typing in the diagnosis and treatment planning of patients with metastatic cancer with unknown or uncertain diagnosis.

A Novel Breast Cancer Index for Prediction of Distant Recurrence in HR+ Early-Stage Breast Cancer with One to Three Positive Nodes.

Purpose: The study objective was to characterize the prognostic performance of a novel Breast Cancer Index model (BCIN+), an integration of BCI gene expression, tumor size, and grade, specifically developed for assessment of distant recurrence (DR) risk in HR+ breast cancer patients with one to three positive lymph nodes (pN1).Experimental Design: Analysis was conducted in a well-annotated retrospective series of pN1 patients (N = 402) treated with adjuvant endocrine therapy with or without chemotherapy using a prespecified model. The primary endpoint was time-to-DR. Results were determined blinded to clinical outcome. Kaplan-Meier estimates of overall (0-15 years) and late (≥5 years) DR, HRs, and 95% confidence interval (CIs) were estimated. Likelihood ratio statistics assessed relative contributions of prognostic information.Results: BCIN+ classified 81 patients (20%) as low risk with a 15-year DR rate of 1.3% (95% CI, 0.0%-3.7%) versus 321 patients as high risk with a DR rate of 29.0% (95% CI, 23.2%-34.4%). In patients DR-free for ≥5 years (n = 349), the late DR rate was 1.3% (95% CI, 0.0%-3.7%) and 16.1% (95% CI, 10.6%-21.3%) in low- and high-risk groups, respectively. BCI gene expression alone was significantly prognostic (ΔLR-χ2 = 20.12; P < 0.0001). Addition of tumor size (ΔLR-χ2 = 13.29, P = 0.0003) and grade (ΔLR-χ2 = 12.72; P = 0.0004) significantly improved prognostic performance. BCI added significant prognostic information to tumor size (ΔLR-χ2 = 17.55; P < 0.0001); addition to tumor grade was incremental (ΔLR-χ2 = 2.38; P = 0.1) with considerable overlap between prognostic values (ΔLR-χ2 = 17.74).Conclusions: The integrated BCIN+ identified 20% of pN1 patients with limited risk of recurrence over 15 years, in whom extended endocrine treatment may be spared. Ongoing studies will characterize combined clinical-genomic risk assessment in node-positive patients. Clin Cancer Res; 23(23); 7217-24. ©2017 AACR.

Risk stratification with Breast Cancer Index for late distant recurrence in patients with clinically low-risk (T1N0) estrogen receptor-positive breast cancer.

Patients with early-stage, hormone receptor-positive breast cancer with favorable clinicopathologic features are often not recommended for extended endocrine therapy. However, even patients with T1N0 disease remain at significant risk of distant recurrence up to 15 years following 5 years of endocrine therapy, highlighting the need for further stratification based on individualized risk to select patients for extended endocrine therapy. In this study, the incremental utility of genomic classification to stratify clinically low-risk patients for late distant recurrence was evaluated using the Breast Cancer Index. In 547 T1N0 patients from two cohorts that were disease-free at 5 years post-diagnosis, Breast Cancer Index categorized 32 and 36% from each cohort, respectively, with high risk of late distant recurrence that was associated with significantly reduced distant recurrence-free survival (86.7 and 89.6%) between years 5-15 and 5-10 compared to Breast Cancer Index low risk (95.4%; P = 0.0263 and 98.4%; P = 0.008). Findings support consideration of genomic classification in clinically low-risk hormone receptor-positive patients to identify candidates for extended endocrine therapy.

Cross-Stratification and Differential Risk by Breast Cancer Index and Recurrence Score in Women with Hormone Receptor-Positive Lymph Node-Negative Early-Stage Breast Cancer.

PURPOSE: Previous results from the TransATAC study demonstrated that both the Breast Cancer Index (BCI) and the OncotypeDX Recurrence Score (RS) added significant prognostic information to clinicopathologic factors over a 10-year period. Here, we examined cross-stratification between BCI and RS to directly compare their prognostic accuracy at the individual patient level. EXPERIMENTAL DESIGN: A total of 665 patients with hormone receptor-positive (HR+) and lymph node-negative disease were included in this retrospective analysis. BCI and RS risk groups were determined using predefined clinical cut-off points. Kaplan-Meier estimates of 10-year risk of distant recurrence (DR) and log-rank tests were used to examine cross-stratification between BCI and RS. RESULTS: As previously reported, both RS and BCI were significantly prognostic in years 0 to 10. BCI provided significant additional prognostic information to the Clinical Treatment Score (CTS) plus RS (ΔLR-χ2 = 11.09; P < 0.001), whereas no additional prognostic information was provided by RS to CTS plus BCI (ΔLR-χ2 = 2.22; P = 0.1). Restratification by BCI of the low and intermediate RS risk groups led to subgroups with significantly different DR rates (P < 0.001 and P = 0.003, respectively). In contrast, restratified subgroups created by RS of BCI risk groups did not differ significantly. CONCLUSIONS: In this retrospective analysis, BCI demonstrated increased prognostic accuracy versus RS. Notably, BCI identified subsets of RS low and RS intermediate risk patients with significant and clinically relevant rates of DR. These results indicate that additional subsets of women with HR+, lymph node-negative breast cancer identified by BCI may be suitable candidates for adjuvant chemotherapy or extended endocrine therapy. Clin Cancer Res; 22(20); 5043-8. ©2016 AACRSee related commentary by Brufsky and Davidson, p. 4963.

Molecular classification of cancer with the 92-gene assay in cytology and limited tissue samples.

BACKGROUND: Detailed molecular evaluation of cytology and limited tissue samples is increasingly becoming the standard for cancer care. Reproducible and accurate diagnostic approaches with reduced demands on cellularity are an ongoing unmet need. This study evaluated the performance of a 92-gene assay for molecular diagnosis of tumor type/subtype in cytology and limited tissue samples. METHODS: Clinical validation of accuracy for the 92-gene assay in limited tissue samples such as cytology cell blocks, core biopsies and small excisions was conducted in a blinded multi-institutional study (N = 109, 48% metastatic, 53% grade II and III). Analytical success rate and diagnostic utility were evaluated in a consecutive series of 644 cytology cases submitted for clinical testing. RESULTS: The 92-gene assay demonstrated 91% sensitivity (95% CI [0.84, 0.95]) for tumor classification, with high accuracy maintained irrespective of specimen type (100%, 92%, and 86% in FNA/cytology cell blocks, core biopsies, and small excisions, respectively; p = 0.26). The assay performed equally well for metastatic versus primary tumors (90% vs 93%, p = 0.73), and across histologic grades (100%, 90%, 89%, in grades I, II, and III, respectively; p = 0.75). In the clinical case series, a molecular diagnosis was reported in 87% of the 644 samples, identifying 23 different tumor types and allowing for additional mutational analysis in selected cases. CONCLUSIONS: These findings demonstrate high accuracy and analytical success rate of the 92-gene assay, supporting its utility in the molecular diagnosis of cancer for specimens with limited tissue.

Prospective assessment of the decision-making impact of the Breast Cancer Index in recommending extended adjuvant endocrine therapy for patients with early-stage ER-positive breast cancer.

Extended adjuvant endocrine therapy (10 vs. 5 years) trials have demonstrated improved outcomes in early-stage estrogen receptor (ER)-positive breast cancer; however, the absolute benefit is modest, and toxicity and tolerability challenges remain. Predictive and prognostic information from genomic analysis may help inform this clinical decision. The purpose of this study was to assess the impact of the Breast Cancer Index (BCI) on physician recommendations for extended endocrine therapy and on patient anxiety and decision conflict. Patients with stage I-III, ER-positive breast cancer who completed at least 3.5 years of adjuvant endocrine therapy were offered participation. Genomic classification with BCI was performed on archived tumor tissues and the results were reported to the treating physician who discussed results with the patient. Patients and physicians completed pre- and post-test questionnaires regarding preferences for extended endocrine therapy. Patients also completed the validated traditional Decisional Conflict Scale (DCS) and State Trait Anxiety Inventory forms (STAI-Y1) pre- and post-test. 96 patients were enrolled at the Yale Cancer Center [median age 60.5 years (range 45-87), 79% postmenopausal, 60% stage I). BCI predicted a low risk of late recurrence in 59% of patients versus intermediate/high in 24 and 17%, respectively. Physician recommendations for extended endocrine therapy changed for 26% of patients after considering BCI results, with a net decrease in recommendations for extended endocrine therapy from 74 to 54%. After testing, fewer patients wanted to continue extended therapy and decision conflict and anxiety also decreased. Mean STAI and DCS scores were 31.3 versus 29.1 (p = 0.031) and 20.9 versus 10.8 (p < 0.001) pre- and post-test, respectively. Incorporation of BCI into risk/benefit discussions regarding extended endocrine therapy resulted in changes in treatment recommendations and improved patient satisfaction.

Health economic analysis of Breast Cancer Index in patients with ER+, LN- breast cancer.

OBJECTIVES: Breast Cancer Index (BCI) is a novel gene expression-based test for patients with estrogen receptor positive (ER+), lymph node negative (LN-) breast cancer that predicts risk of recurrence over 10 years, and also specifically predicts risk of late (≥5 y) recurrences and likelihood of benefit from extended (≥5 y) endocrine therapy. The objective of this study was to evaluate cost utility of BCI from a US third-party payer perspective. STUDY DESIGN: Two fact-based economic models were developed to project the cost and effectiveness of BCI in a hypothetical population of patients with ER+, LN- breast cancer compared with standard clinicopathologic diagnostic modalities. METHODS: Costs associated with adjuvant chemotherapy, toxicity, followup, endocrine therapy, and recurrence were modeled over 10 years. The models examined cost utility compared with standard practice when used at diagnosis and in patients disease-free at 5 years post diagnosis. RESULTS: Use of BCI was projected to be cost saving in both models. In the newly diagnosed population, net cost savings were $3803 per patient tested. In the 5 years post diagnosis population, BCI was projected to yield a net cost savings of $1803 per patient tested. Sensitivity analyses demonstrated that BCI was cost saving across a wide range of clinically relevant input assumptions. CONCLUSIONS: BCI was projected to be cost saving when used either at diagnosis or at 5 years post diagnosis. Cost savings are achieved through projected impact on adjuvant chemotherapy use, extended endocrine therapy use, and endocrine therapy compliance. These findings require validation in additional cohorts, including studies of real-world clinical practice.

Cost effectiveness of a 92-gene assay for the diagnosis of metastatic cancer.

OBJECTIVES: To estimate the clinical and economic trade-offs involved in using a molecular assay (92-gene assay, CancerTYPE ID) to aid in identifying the primary site of difficult-to-diagnose metastatic cancers and to explore whether the 92-gene assay can be used to standardize the diagnostic process and costs for clinicians, patients, and payers. METHODS: Four decision-analytic models were developed to project the lifetime clinical and economic impact of incorporating the 92-gene assay compared with standard care alone. For each model, total and incremental costs, life-years, quality-adjusted life-years (QALYs), incremental cost-effectiveness ratios (ICERs), and the proportion of patients treated correctly versus incorrectly were projected from the payer perspective. Model inputs were based on published literature, analyses of SEER (Surveillance Epidemiology and End RESULTS) data, publicly available data, and interviews with clinical experts. RESULTS: In all four models, the 92-gene assay increased the proportion of patients treated correctly, decreased the proportion of patients treated with empiric therapy, and increased quality-adjusted survival. In the primary model, the ICER was $50,273/QALY; thus, the 92-gene assay is therefore cost effective when considering a societal willingness-to-pay threshold of $100,000/QALY. These findings were robust across sensitivity analyses. CONCLUSIONS: Use of the 92-gene assay for diagnosing metastatic tumors of uncertain origin is associated with reduced misdiagnoses, increased survival, and improved quality of life. Incorporating the assay into current practice is a cost-effective approach to standardizing diagnostic methods while improving patient care. Limitations of this analysis are the lack of data availability and resulting modeling simplifications, although sensitivity analyses showed these to not be key drivers of results.

Prediction of late distant recurrence in patients with oestrogen-receptor-positive breast cancer: a prospective comparison of the breast-cancer index (BCI) assay, 21-gene recurrence score, and IHC4 in the TransATAC study population.

BACKGROUND: Biomarkers to improve the risk-benefit of extended adjuvant endocrine therapy for late recurrence in patients with oestrogen-receptor-positive breast cancer would be clinically valuable. We compared the prognostic ability of the breast-cancer index (BCI) assay, 21-gene recurrence score (Oncotype DX), and an immunohistochemical prognostic model (IHC4) for both early and late recurrence in patients with oestrogen-receptor-positive, node-negative (N0) disease who took part in the Arimidex, Tamoxifen, Alone or in Combination (ATAC) clinical trial. METHODS: In this prospective comparison study, we obtained archival tumour blocks from the TransATAC tissue bank from all postmenopausal patients with oestrogen-receptor-positive breast cancer from whom the 21-gene recurrence score and IHC4 values had already been derived. We did BCI analysis in matched samples with sufficient residual RNA using two BCI models-cubic (BCI-C) and linear (BCI-L)-using previously validated cutoffs. We assessed prognostic ability of BCI for distant recurrence over 10 years (the primary endpoint) and compared it with that of the 21-gene recurrence score and IHC4. We also tested the ability of the assays to predict early (0-5 years) and late (5-10 years) distant recurrence. To assess the ability of the biomarkers to predict recurrence beyond standard clinicopathological variables, we calculated the change in the likelihood-ratio χ(2) (LR-Δχ(2)) from Cox proportional hazards models. FINDINGS: Suitable tissue was available from 665 patients with oestrogen-receptor-positive, N0 breast cancer for BCI analysis. The primary analysis showed significant differences in risk of distant recurrence over 10 years in the categorical BCI-C risk groups (p<0·0001) with 6·8% (95% CI 4·4-10·0) of patients in the low-risk group, 17·3% (12·0-24·7) in the intermediate group, and 22·2% (15·3-31·5) in the high-risk group having distant recurrence. The secondary analysis showed that BCI-L was a much stronger predictor for overall (0-10 year) distant recurrence compared with BCI-C (interquartile HR 2·30 [95% CI 1·62-3·27]; LR-Δχ(2)=22·69; p<0·0001). When compared with BCI-L, the 21-gene recurrence score was less predictive (HR 1·48 [95% CI 1·22-1·78]; LR-Δχ(2)=13·68; p=0·0002) and IHC4 was similar (HR 1·69 [95% CI 1·51-2·56]; LR-Δχ(2)=22·83; p<0·0001). All further analyses were done with the BCI-L model. In a multivariable analysis, all assays had significant prognostic ability for early distant recurrence (BCI-L HR 2·77 [95% CI 1·63-4·70], LR-Δχ(2)=15·42, p<0·0001; 21-gene recurrence score HR 1·80 [1·42-2·29], LR-Δχ(2)=18·48, p<0·0001; IHC4 HR 2·90 [2·01-4·18], LR-Δχ(2)=29·14, p<0·0001); however, only BCI-L was significant for late distant recurrence (BCI-L HR 1·95 [95% CI 1·22-3·14], LR-Δχ(2)=7·97, p=0·0048; 21-gene recurrence score HR 1·13 [0·82-1·56], LR-Δχ(2)=0·48, p=0·47; IHC4 HR 1·30 [0·88-1·94], LR-Δχ(2)=1·59, p=0·20). INTERPRETATION: BCI-L was the only significant prognostic test for risk of both early and late distant recurrence and identified two risk populations for each timeframe. It could help to identify patients at high risk for late distant recurrence who might benefit from extended endocrine or other therapy. FUNDING: Avon Foundation, National Institutes of Health, Breast Cancer Foundation, US Department of Defense Breast Cancer Research Program, Susan G Komen for the Cure, Breakthrough Breast Cancer through the Mary-Jean Mitchell Green Foundation, AstraZeneca, Cancer Research UK, and the National Institute for Health Research Biomedical Research Centre at the Royal Marsden (London, UK).