Computational limits to the legibility of the imaged human brain.

Our knowledge of the organisation of the human brain at the population-level is yet to translate into power to predict functional differences at the individual-level, limiting clinical applications and casting doubt on the generalisability of inferred mechanisms. It remains unknown whether the difficulty arises from the absence of individuating biological patterns within the brain, or from limited power to access them with the models and compute at our disposal. Here we comprehensively investigate the resolvability of such patterns with data and compute at unprecedented scale. Across 23 810 unique participants from UK Biobank, we systematically evaluate the predictability of 25 individual biological characteristics, from all available combinations of structural and functional neuroimaging data. Over 4526 GPU*hours of computation, we train, optimize, and evaluate out-of-sample 700 individual predictive models, including fully-connected feed-forward neural networks of demographic, psychological, serological, chronic disease, and functional connectivity characteristics, and both uni- and multi-modal 3D convolutional neural network models of macro- and micro-structural brain imaging. We find a marked discrepancy between the high predictability of sex (balanced accuracy 99.7%), age (mean absolute error 2.048 years, R2 0.859), and weight (mean absolute error 2.609Kg, R2 0.625), for which we set new state-of-the-art performance, and the surprisingly low predictability of other characteristics. Neither structural nor functional imaging predicted an individual's psychology better than the coincidence of common chronic disease (p < 0.05). Serology predicted chronic disease (p < 0.05) and was best predicted by it (p < 0.001), followed by structural neuroimaging (p < 0.05). Our findings suggest either more informative imaging or more powerful models will be needed to decipher individual level characteristics from the human brain. We make our models and code openly available.

Compressed representation of brain genetic transcription.

The architecture of the brain is too complex to be intuitively surveyable without the use of compressed representations that project its variation into a compact, navigable space. The task is especially challenging with high-dimensional data, such as gene expression, where the joint complexity of anatomical and transcriptional patterns demands maximum compression. The established practice is to use standard principal component analysis (PCA), whose computational felicity is offset by limited expressivity, especially at great compression ratios. Employing whole-brain, voxel-wise Allen Brain Atlas transcription data, here we systematically compare compressed representations based on the most widely supported linear and non-linear methods-PCA, kernel PCA, non-negative matrix factorisation (NMF), t-stochastic neighbour embedding (t-SNE), uniform manifold approximation and projection (UMAP), and deep auto-encoding-quantifying reconstruction fidelity, anatomical coherence, and predictive utility across signalling, microstructural, and metabolic targets, drawn from large-scale open-source MRI and PET data. We show that deep auto-encoders yield superior representations across all metrics of performance and target domains, supporting their use as the reference standard for representing transcription patterns in the human brain.

Fatigue and endocrine symptoms among women with early breast cancer randomized to endocrine versus chemoendocrine therapy: Results from the TAILORx patient-reported outcomes substudy.

BACKGROUND: TAILORx (Trial Assigning Individualized Options for Treatment) prospectively assessed fatigue and endocrine symptoms among women with early-stage hormone receptor-positive breast cancer and a midrange risk of recurrence who were randomized to endocrine therapy (E) or chemotherapy followed by endocrine therapy (CT+E). METHODS: Participants completed the Functional Assessment of Chronic Illness Therapy-Fatigue, the Patient-Reported Outcomes Measurement Information System-Fatigue Short Form, and the Functional Assessment of Cancer Therapy-Endocrine Symptoms at the baseline and at 3, 6, 12, 24, and 36 months. Linear regression was used to model outcomes on baseline symptoms, treatment, and other factors. RESULTS: Participants (n = 458) in both treatment arms reported greater fatigue and endocrine symptoms at early follow-up in comparison with the baseline. The magnitude of change in fatigue was significantly greater for the CT+E arm than the E arm at 3 and 6 months but not at 12, 24, or 36 months. The CT+E arm reported significantly greater changes in endocrine symptoms from the baseline to 3 months in comparison with the E arm; change scores were not significantly different at later time points. Endocrine symptom trajectories by treatment differed by menopausal status, with the effect larger and increasing for postmenopausal patients. CONCLUSIONS: Adjuvant CT+E was associated with greater increases in fatigue and endocrine symptoms at early time points in comparison with E. These differences lessened over time, and this demonstrated early chemotherapy effects more than long-term ones. Treatment arm differences in endocrine symptoms were more evident in postmenopausal patients. LAY SUMMARY: Participants in TAILORx (Trial Assigning Individualized Options for Treatment) with early-stage hormone receptor-positive breast cancer and an intermediate risk of recurrence were randomly assigned to endocrine or chemoendocrine therapy. Four hundred fifty-eight women reported fatigue and endocrine symptoms at the baseline and at 3, 6, 12, 24, and 36 months. Both groups reported greater symptoms at early follow-up versus the baseline. Increases in fatigue were greater for the chemoendocrine group than the endocrine group at 3 and 6 months but not later. The chemoendocrine group reported greater changes in endocrine symptoms in comparison with the endocrine group at 3 months but not later.

Clinical and Genomic Risk to Guide the Use of Adjuvant Therapy for Breast Cancer.

BACKGROUND: The use of adjuvant chemotherapy in patients with breast cancer may be guided by clinicopathological factors and a score based on a 21-gene assay to determine the risk of recurrence. Whether the level of clinical risk of breast cancer recurrence adds prognostic information to the recurrence score is not known. METHODS: We performed a prospective trial involving 9427 women with hormone-receptor-positive, human epidermal growth factor receptor 2-negative, axillary node-negative breast cancer, in whom an assay of 21 genes had been performed, and we classified the clinical risk of recurrence of breast cancer as low or high on the basis of the tumor size and histologic grade. The effect of clinical risk was evaluated by calculating hazard ratios for distant recurrence with the use of Cox proportional-hazards models. The initial endocrine therapy was tamoxifen alone in the majority of the premenopausal women who were 50 years of age or younger. RESULTS: The level of clinical risk was prognostic of distant recurrence in women with an intermediate 21-gene recurrence score of 11 to 25 (on a scale of 0 to 100, with higher scores indicating a worse prognosis or a greater potential benefit from chemotherapy) who were randomly assigned to endocrine therapy (hazard ratio for the comparison of high vs. low clinical risk, 2.73; 95% confidence interval [CI], 1.93 to 3.87) or to chemotherapy plus endocrine (chemoendocrine) therapy (hazard ratio, 2.41; 95% CI, 1.66 to 3.48) and in women with a high recurrence score (a score of 26 to 100), all of whom were assigned to chemoendocrine therapy (hazard ratio, 3.17; 95% CI, 1.94 to 5.19). Among women who were 50 years of age or younger who had received endocrine therapy alone, the estimated (±SE) rate of distant recurrence at 9 years was less than 5% (≤1.8±0.9%) with a low recurrence score (a score of 0 to 10), irrespective of clinical risk, and 4.7±1.0% with an intermediate recurrence score and low clinical risk. In this age group, the estimated distant recurrence at 9 years exceeded 10% among women with a high clinical risk and an intermediate recurrence score who received endocrine therapy alone (12.3±2.4%) and among those with a high recurrence score who received chemoendocrine therapy (15.2±3.3%). CONCLUSIONS: Clinical-risk stratification provided prognostic information that, when added to the 21-gene recurrence score, could be used to identify premenopausal women who could benefit from more effective therapy. (Funded by the National Cancer Institute and others; ClinicalTrials.gov number, NCT00310180.).

Breast cancer patients' insurance status and residence zip code correlate with early discontinuation of endocrine therapy: An analysis of the ECOG-ACRIN TAILORx trial.

BACKGROUND: Early discontinuation is a substantial barrier to the delivery of endocrine therapies (ETs) and may influence recurrence and survival. The authors investigated the association between early discontinuation of ET and social determinants of health, including insurance coverage and the neighborhood deprivation index (NDI), which was measured on the basis of patients' zip codes, in breast cancer. METHODS: In this retrospective analysis of a prospective randomized clinical trial (Trial Assigning Individualized Options for Treatment), women with hormone receptor-positive, human epidermal growth factor receptor 2-negative breast cancer who started ET within a year of study entry were included. Early discontinuation was calculated as stopping ET within 4 years of its start for reasons other than distant recurrence or death via Kaplan-Meier estimates. A Cox proportional hazards joint model was used to analyze the association between early discontinuation of ET and factors such as the study-entry insurance and NDI, with adjustments made for other variables. RESULTS: Of the included 9475 women (mean age, 55.6 years; White race, 84%), 58.0% had private insurance, whereas 11.7% had Medicare, 5.8% had Medicaid, 3.8% were self-pay, and 19.1% were treated at international sites. The early discontinuation rate was 12.3%. Compared with those with private insurance, patients with Medicaid (hazard ratio [HR], 1.53; 95% confidence interval [CI], 1.23-1.92) and self-pay patients (HR, 1.65; 95% CI, 1.25-2.17) had higher early discontinuation. Participants with a first-quartile NDI (highest deprivation) had a higher probability of discontinuation than those with a fourth-quartile NDI (lowest deprivation; HR, 1.34; 95% CI, 1.11-1.62). CONCLUSIONS: Patients' insurance and zip code at study entry play roles in adherence to ET, with uninsured and underinsured patients having a high rate of treatment nonadherence. Early identification of patients at risk may improve adherence to therapy. LAY SUMMARY: In this retrospective analysis of 9475 women with breast cancer participating in a clinical trial (Trial Assigning Individualized Options for Treatment), Medicaid and self-pay patients (compared with those with private insurance) and those in the highest quartile of neighborhood deprivation scores (compared with those in the lowest quartile) had a higher probability of early discontinuation of endocrine therapy. These social determinants of health assume larger importance with the expected increase in unemployment rates and loss of insurance coverage in the aftermath of the coronavirus disease 2019 pandemic. Early identification of patients at risk and enrollment in insurance optimization programs may improve the persistence of therapy.

Tolerability of bevacizumab and chemotherapy in a phase 3 clinical trial with human epidermal growth factor receptor 2-negative breast cancer: A trajectory analysis of adverse events.

BACKGROUND: E5103 was a study designed to evaluate the efficacy and safety of bevacizumab. It was a negative trial for the end points of invasive disease-free survival and overall survival. The current work examines the tolerability of bevacizumab and other medication exposures with respect to clinical outcomes and patient-reported outcomes (PROs). METHODS: Adverse events (AEs) collected from the Common Terminology Criteria for Adverse Events were summarized to form an AE profile at each treatment cycle. All-grade and high-grade events were separately analyzed. The change in the AE profile over the treatment cycle was delineated as distinct AE trajectory clusters. AE-related and any-reason early treatment discontinuations were treated as clinical outcome measures. PROs were measured with the Functional Assessment of Cancer Therapy-Breast + Lymphedema. The relationships between the AE trajectory and early treatment discontinuation as well as PROs were analyzed. RESULTS: More than half of all AEs (57.5%) were low-grade. A cluster of patients with broad and mixed AE (all-grade) trajectory grades was significantly associated with any-reason early treatment discontinuation (odds ratio [OR], 2.87; P = .01) as well as AE-related discontinuation (OR, 4.14; P = .001). This cluster had the highest count of all-grade AEs per cycle in comparison with other clusters. Another cluster of patients with primary neuropathic AEs in their trajectories had poorer physical well-being in comparison with a trajectory of no or few AEs (P < .01). A high-grade AE trajectory did not predict discontinuations. CONCLUSIONS: A sustained and cumulative burden of across-the-board toxicities, which were not necessarily all recognized as high-grade AEs, contributed to early treatment discontinuation. Patients with neuropathic all-grade AEs may require additional attention for preventing deterioration in their physical well-being.

Adjuvant Chemotherapy Guided by a 21-Gene Expression Assay in Breast Cancer.

BACKGROUND: The recurrence score based on the 21-gene breast cancer assay predicts chemotherapy benefit if it is high and a low risk of recurrence in the absence of chemotherapy if it is low; however, there is uncertainty about the benefit of chemotherapy for most patients, who have a midrange score. METHODS: We performed a prospective trial involving 10,273 women with hormone-receptor-positive, human epidermal growth factor receptor 2 (HER2)-negative, axillary node-negative breast cancer. Of the 9719 eligible patients with follow-up information, 6711 (69%) had a midrange recurrence score of 11 to 25 and were randomly assigned to receive either chemoendocrine therapy or endocrine therapy alone. The trial was designed to show noninferiority of endocrine therapy alone for invasive disease-free survival (defined as freedom from invasive disease recurrence, second primary cancer, or death). RESULTS: Endocrine therapy was noninferior to chemoendocrine therapy in the analysis of invasive disease-free survival (hazard ratio for invasive disease recurrence, second primary cancer, or death [endocrine vs. chemoendocrine therapy], 1.08; 95% confidence interval, 0.94 to 1.24; P=0.26). At 9 years, the two treatment groups had similar rates of invasive disease-free survival (83.3% in the endocrine-therapy group and 84.3% in the chemoendocrine-therapy group), freedom from disease recurrence at a distant site (94.5% and 95.0%) or at a distant or local-regional site (92.2% and 92.9%), and overall survival (93.9% and 93.8%). The chemotherapy benefit for invasive disease-free survival varied with the combination of recurrence score and age (P=0.004), with some benefit of chemotherapy found in women 50 years of age or younger with a recurrence score of 16 to 25. CONCLUSIONS: Adjuvant endocrine therapy and chemoendocrine therapy had similar efficacy in women with hormone-receptor-positive, HER2-negative, axillary node-negative breast cancer who had a midrange 21-gene recurrence score, although some benefit of chemotherapy was found in some women 50 years of age or younger. (Funded by the National Cancer Institute and others; TAILORx ClinicalTrials.gov number, NCT00310180 .).

Noninferiority trials with nonadherence to the assigned randomized treatment.

BACKGROUND: Nonadherence to treatment assignment in a noninferiority randomized trial is especially problematic because it attenuates observed differences between the treatment arms, possibly leading one to conclude erroneously that a truly inferior experimental therapy is noninferior to a standard therapy (inflated type 1 error probability). The Lachin-Foulkes adjustment is an increase in the sample size to account for random nonadherence for the design of a superiority trial with a time-to-event outcome; it has not been explored in the noninferiority trial setting nor with nonrandom nonadherence. Noninferiority trials where patients have knowledge of a personal prognostic risk score may lead to nonrandom nonadherence, as patients with a relatively high risk may be more likely to not adhere to the random assignment to the (reduced) experimental therapy, and patients with a relatively low risk score may be more likely to not adhere to the random assignment to the (more aggressive) standard therapy. METHODS: We investigated via simulations the properties of the Lachin-Foulkes adjustment in the noninferiority setting. We considered nonrandom in addition to random nonadherence to the treatment assignment. For nonrandom nonadherence, we used the scenario where a risk score, potentially associated with the between-arm treatment difference, influences patients' nonadherence. A sensitivity analysis is proposed for addressing the nonrandom nonadherence for this scenario. The noninferiority TAILORx adjuvant breast cancer trial, where eligibility was based on a genomic risk score, is used as an example throughout. RESULTS: The Lachin-Foulkes adjustment to the sample size improves the operating characteristics of noninferiority trials with random nonadherence. However, to maintain type 1 error probability, it is critical to adjust the noninferiorty margin as well as the sample size. With nonrandom nonadherence that is associated with a prognostic risk score, the type 1 error probability of the Lachin-Foulkes adjustment can be inflated (e.g. doubled) when the nonadherence is larger in the experimental arm than the standard arm. The proposed sensitivity analysis lessens the inflation in this situation. CONCLUSION: The Lachin-Foulkes adjustment to the sample size and noninferiority margin is a useful simple technique for attenuating the effects of random nonadherence in the noninferiority setting. With nonrandom nonadherence associated with a risk score known to the patients, the type 1 error probability can be inflated in certain situations. A proposed sensitivity analysis for these situations can attenuate the inflation.

Image-Based Comparison Between the Bilateral Symmetry of the Distal Radii Through Established Measures.

PURPOSE: Anthropometric assessment of bony structures in the body is important for preoperative computer-aided surgery, implant design, finite element modeling, and biomechanical studies investigating joint structure and function. The use of the contralateral limb in surgery and clinical practice relies on the assumption that the right and left limbs of an individual are symmetric. Therefore, the purpose of this study was to quantify the bilateral symmetry of the bony structures of the distal radius using 3-dimensional (3D) computed tomography. METHODS: We collected computed tomography images of 37 paired, fresh-frozen, healthy cadaveric male upper limbs (aged 75.4 ± 8.3 years). Three-dimensional reconstructed models were created using semiautomatic segmentation. Using the 3D models, we measured 3D radial inclination, 3D volar tilt, 3D radial height, medial volar cortical angle, middle volar cortical angle, and lateral volar cortical angle and compared them between sides. RESULTS: There were no statistically significant differences measured between right and left distal radius in 37 paired wrists. Mean radial height was 12.81 mm (SD, 1.74 mm) on the left and 12.88 mm (SD, 1.72 mm) on the right. Mean volar tilt was 10.74° (SD, 3.74°) and 10.77° (SD, 3.19°) and radial inclination was 24.05° (SD, 2.63°) and 24.18° (SD, 3.41°) on the left and right, respectively. Mean volar cortical angle across the radius was 140.9° (SD, 7.9°) on the left and 140.1° (SD, 7.9°) on the right. CONCLUSIONS: Direct bilateral comparison of the distal radius and wrist joints is useful to predict normal anatomy of the injured radius, because bilateral similarities exist. CLINICAL RELEVANCE: This article provides a comprehensive list of measurements of the distal radius compared bilaterally using a 3D model. From this study, we found that the contralateral radius can be used as a benchmark with which to compare fracture reduction and to manage malunions during the preoperative planning of corrective osteotomies. It can also be used to define normal anatomy.

Prospective Validation of a 21-Gene Expression Assay in Breast Cancer.

BACKGROUND: Prior studies with the use of a prospective-retrospective design including archival tumor samples have shown that gene-expression assays provide clinically useful prognostic information. However, a prospectively conducted study in a uniformly treated population provides the highest level of evidence supporting the clinical validity and usefulness of a biomarker. METHODS: We performed a prospective trial involving women with hormone-receptor-positive, human epidermal growth factor receptor type 2 (HER2)-negative, axillary node-negative breast cancer with tumors of 1.1 to 5.0 cm in the greatest dimension (or 0.6 to 1.0 cm in the greatest dimension and intermediate or high tumor grade) who met established guidelines for the consideration of adjuvant chemotherapy on the basis of clinicopathologic features. A reverse-transcriptase-polymerase-chain-reaction assay of 21 genes was performed on the paraffin-embedded tumor tissue, and the results were used to calculate a score indicating the risk of breast-cancer recurrence; patients were assigned to receive endocrine therapy without chemotherapy if they had a recurrence score of 0 to 10, indicating a very low risk of recurrence (on a scale of 0 to 100, with higher scores indicating a greater risk of recurrence). RESULTS: Of the 10,253 eligible women enrolled, 1626 women (15.9%) who had a recurrence score of 0 to 10 were assigned to receive endocrine therapy alone without chemotherapy. At 5 years, in this patient population, the rate of invasive disease-free survival was 93.8% (95% confidence interval [CI], 92.4 to 94.9), the rate of freedom from recurrence of breast cancer at a distant site was 99.3% (95% CI, 98.7 to 99.6), the rate of freedom from recurrence of breast cancer at a distant or local-regional site was 98.7% (95% CI, 97.9 to 99.2), and the rate of overall survival was 98.0% (95% CI, 97.1 to 98.6). CONCLUSIONS: Among patients with hormone-receptor-positive, HER2-negative, axillary node-negative breast cancer who met established guidelines for the recommendation of adjuvant chemotherapy on the basis of clinicopathologic features, those with tumors that had a favorable gene-expression profile had very low rates of recurrence at 5 years with endocrine therapy alone. (Funded by the National Cancer Institute and others; ClinicalTrials.gov number, NCT00310180.).

Estimating treatment effect in a proportional hazards model in randomized clinical trials with all-or-nothing compliance.

We consider methods for estimating the treatment effect and/or the covariate by treatment interaction effect in a randomized clinical trial under noncompliance with time-to-event outcome. As in Cuzick et al. (2007), assuming that the patient population consists of three (possibly latent) subgroups based on treatment preference: the ambivalent group, the insisters, and the refusers, we estimate the effects among the ambivalent group. The parameters have causal interpretations under standard assumptions. The article contains two main contributions. First, we propose a weighted per-protocol (Wtd PP) estimator through incorporating time-varying weights in a proportional hazards model. In the second part of the article, under the model considered in Cuzick et al. (2007), we propose an EM algorithm to maximize a full likelihood (FL) as well as the pseudo likelihood (PL) considered in Cuzick et al. (2007). The E step of the algorithm involves computing the conditional expectation of a linear function of the latent membership, and the main advantage of the EM algorithm is that the risk parameters can be updated by fitting a weighted Cox model using standard software and the baseline hazard can be updated using closed-form solutions. Simulations show that the EM algorithm is computationally much more efficient than directly maximizing the observed likelihood. The main advantage of the Wtd PP approach is that it is more robust to model misspecifications among the insisters and refusers since the outcome model does not impose distributional assumptions among these two groups.

Features of lymphatic dysfunction in compressed skin tissues - Implications in pressure ulcer aetiology.

OBJECTIVE: Impaired lymph formation and clearance has previously been proposed as a contributory factor in the development of pressure ulcers. The present study has been designed to trial fluorescence lymphangiography for establishing how lymphatic function is altered under a clinically relevant form of mechanical loading. METHODS: Lymph formation and clearance was traced in both forearms by an intradermal injection of indocyanine green (ICG) (50 µl, 0.05%w/v), imaged using a commercial near-infrared fluorescence imaging unit (Fluobeam(®) 800). External uniaxial loading equivalent to a pressure of 60 mmHg was applied for 45 min in one arm using a custom-built indenter. RESULTS: Loading was associated with a decreased frequency of normal directional drainage (DD) of ICG within delineated vessels, both immediately after loading and 45 min thereafter. Loading was also associated with non-directional drainage (NDD) of ICG within the interstitium. Signal intensity within NDD was often greatest at areas of stress concentration, producing a 'halo pattern', corresponding to the rounded edges of the indenter. CONCLUSIONS: These results suggest that loading skin with a clinically relevant magnitude of pressure alters both lymph formation and clearance. Further work to quantify impaired clearance under mechanical loading could provide valuable insight into their involvement in the development of pressure ulcers.

The minimal computational substrate of fluid intelligence.

The quantification of cognitive powers rests on identifying a behavioural task that depends on them. Such dependence cannot be assured, for the powers a task invokes cannot be experimentally controlled or constrained a priori, resulting in unknown vulnerability to failure of specificity and generalisability. Evaluating a compact version of Raven's Advanced Progressive Matrices (RAPM), a widely used clinical test of fluid intelligence, we show that LaMa, a self-supervised artificial neural network trained solely on the completion of partially masked images of natural environmental scenes, achieves representative human-level test scores a prima vista, without any task-specific inductive bias or training. Compared with cohorts of healthy and focally lesioned participants, LaMa exhibits human-like variation with item difficulty, and produces errors characteristic of right frontal lobe damage under degradation of its ability to integrate global spatial patterns. LaMa's narrow training and limited capacity suggest matrix-style tests may be open to computationally simple solutions that need not necessarily invoke the substrates of reasoning.

Realistic morphology-preserving generative modelling of the brain.

Medical imaging research is often limited by data scarcity and availability. Governance, privacy concerns and the cost of acquisition all restrict access to medical imaging data, which, compounded by the data-hungry nature of deep learning algorithms, limits progress in the field of healthcare AI. Generative models have recently been used to synthesize photorealistic natural images, presenting a potential solution to the data scarcity problem. But are current generative models synthesizing morphologically correct samples? In this work we present a three-dimensional generative model of the human brain that is trained at the necessary scale to generate diverse, realistic-looking, high-resolution and morphologically preserving samples and conditioned on patient characteristics (for example, age and pathology). We show that the synthetic samples generated by the model preserve biological and disease phenotypes and are realistic enough to permit use downstream in well-established image analysis tools. While the proposed model has broad future applicability, such as anomaly detection and learning under limited data, its generative capabilities can be used to directly mitigate data scarcity, limited data availability and algorithmic fairness.

Clinical and Genomic Risk for Late Breast Cancer Recurrence and Survival.

BACKGROUND: The 21-gene recurrence score (RS) assay (Oncotype DX) is used to guide adjuvant chemotherapy use for patients with hormone receptor-positive, HER2 (human epidermal growth factor receptor 2)-negative, axillary node-negative breast cancer. Its role, however, in providing prognostic information for late distant recurrence when added to clinicopathologic prognostic factors is unknown. METHODS: A patient-specific meta-analysis including 10,004 women enrolled in three trials was updated using extended follow-up data from TAILORx, integrating the RS with histologic grade, tumor size, and age at surgery for the RSClin tool. Cox models integrating clinicopathologic factors and the RS were compared by using likelihood ratio (LR) tests. External validation of prognosis for distant recurrence in years 0 to 10 and 5 to 10 was performed in an independent cohort of 1098 women in a real-world registry. RESULTS: RSClin provided significantly more prognostic information than either the clinicopathologic factors (ΔLR chi-square, 86.2; P<0.001) or RS alone (ΔLR chi-square, 131.0; P<0.001). The model was prognostic in an independent cohort for distant recurrence by 10 years after diagnosis (standardized hazard ratio, 1.56; 95% confidence interval, 1.25 to 1.94), was associated with late distant recurrence risk between 5 and 10 years after diagnosis (standardized hazard ratio, 1.78; 95% confidence interval, 1.25 to 2.55), and approximated the observed 10-year distant recurrence risk (Lin concordance, 0.87) and 5- to 10-year distant recurrence risk (Lin concordance, 0.92). CONCLUSIONS: The 21-gene RS is prognostic for distant recurrence and overall survival in early breast cancer. A model integrating the 21-gene RS and clinicopathologic factors improved estimates of distant recurrence risk compared with either used individually and stratified late distant recurrence risk. (Funded by the National Cancer Institute, National Institutes of Health [U10CA180820, U10CA180794, UG1CA189859, U10CA180868, and U10CA180822] and others.).

Phase II Study of Osimertinib in Patients With Epidermal Growth Factor Receptor Mutations: Results From the NCI-MATCH ECOG-ACRIN (EAY131) Trial Subprotocol E.

PURPOSE: The National Cancer Institute Molecular Analysis for Therapy Choice trial is a signal-finding genomically driven platform trial that assigns patients with any advanced refractory solid tumor, lymphoma, or myeloma to targeted therapies on the basis of next-generation sequencing results. Subprotocol E evaluated osimertinib, an epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor, in patients with EGFR mutations. METHODS: Eligible patients had EGFR mutations (T790M or rare activating) and received osimertinib 80 mg once daily. Patients with lung cancer with EGFR T790M were excluded. The primary end point was objective response rate (ORR), and the secondary end points were 6-month progression-free survival (PFS), overall survival, and toxicity. RESULTS: A total of 19 patients were enrolled: 17 were evaluable for toxicity and 13 for efficacy. The median age of the 13 included in the efficacy analysis was 63 years, 62% had Eastern Cooperative Oncology Group performance status 1, and 31% received >three previous systemic therapies. The most common tumor type was brain cancers (54%). The ORR was 15.4% (n = 2 of 13; 90% CI, 2.8 to 41.0) and 6-month PFS was 16.7% (90% CI, 0 to 34.4). The two confirmed RECIST responses were observed in a patient with neuroendocrine carcinoma not otherwise specified (EGFR exon 20 S768T and exon 18 G719C mutation) and a patient with low-grade epithelial carcinoma of the paranasal sinus (EGFR D770_N771insSVD). The most common (>20%) treatment-related adverse events were diarrhea, thrombocytopenia, and maculopapular rash. CONCLUSION: In this pretreated cohort, osimertinib did not meet the prespecified end point threshold for efficacy, but responses were seen in a neuroendocrine carcinoma with an EGFR exon 20 S768T and exon 18 G719C mutation and an epithelial carcinoma with an EGFR D770_N771insSVD mutation. Osimertinib was well tolerated and had a safety profile consistent with previous studies.

Phase II Trial of Afatinib in Patients With EGFR-Mutated Solid Tumors Excluding Lung Cancer: Results From NCI-MATCH ECOG-ACRIN Trial (EAY131) Subprotocol A.

PURPOSE: National Cancer Institute-Molecular Analysis for Therapy Choice (NCI-MATCH) was a multicohort phase 2 trial that assigned patients with advanced pretreated cancers to molecularly targeted therapies on the basis of tumor genomic testing. NCI-MATCH Arm A evaluated afatinib, an EGFR tyrosine kinase inhibitor (TKI) approved for advanced non-small cell lung cancer, in patients with tumors other than lung cancer harboring EGFR mutations. METHODS: Patients with advanced pretreated cancers other than lung cancer found to have selected actionable EGFR mutations were offered participation in Arm A. Previous therapy with an EGFR TKI was not allowed. Patients received afatinib 40 mg once daily continuously until disease progression or unacceptable toxicity. The primary end point was objective response rate (ORR). Secondary end points included progression-free survival (PFS), 6-month PFS, and overall survival (OS). RESULTS: Seventeen patients received protocol therapy. Tumor types included glioblastoma multiforme (GBM) (13), gliosarcoma (1), adenocarcinoma not otherwise specified (NOS) (2), and adenosquamous carcinoma of the breast (1). Fifty-nine percent of patients received ≥2 lines of previous therapy. The ORR was 11.8% (90% CI, 2.1 to 32.6), with one complete response lasting 16.4 months (GBM harboring a rare exon 18 EGFR-SEPT14 fusion) and one partial response lasting 12.8 months (adenocarcinoma NOS with the classic EGFR mutation, p.Glu746_Ala750del). Three patients had stable disease. The 6-month PFS was 15% (90% CI, 0 to 30.7); the median OS was 9 months (90% CI, 4.6 to 14.0). Rash and diarrhea were the most common toxicities. CONCLUSION: Afatinib had modest activity in a cohort of patients with heavily pretreated cancer with advanced nonlung, EGFR-mutated tumors, but the trial's primary end point was not met. Further evaluation of afatinib in GBM with EGFR exon 18 fusions may be of interest.

Trastuzumab and Pertuzumab in Patients with Non-Breast/Gastroesophageal HER2-Amplified Tumors: Results from the NCI-MATCH ECOG-ACRIN Trial (EAY131) Subprotocol J.

PURPOSE: NCI-MATCH assigned patients with advanced cancer and progression on prior treatment, based on genomic alterations in pretreatment tumor tissue. Arm J (EAY131-J) evaluated the combination of trastuzumab/pertuzumab (HP) across HER2-amplified tumors. PATIENTS AND METHODS: Eligible patients had high levels of HER2 amplification [copy number (CN) ≥7] detected by central next-generation sequencing (NGS) or through NCI-designated laboratories. Patients with breast/gastroesophageal adenocarcinoma and those who received prior HER2-directed therapy were excluded. Enrollment of patients with colorectal cancer was capped at 4 based on emerging data. Patients received HP IV Q3 weeks until progression or unacceptable toxicity. Primary endpoint was objective response rate (ORR); secondary endpoints included progression-free survival (PFS) and overall survival (OS). RESULTS: Thirty-five patients were enrolled, with 25 included in the primary efficacy analysis (CN ≥7 confirmed by a central lab, median CN = 28). Median age was 66 (range, 31-80), and half of all patients had ≥3 prior therapies (range, 1-11). The confirmed ORR was 12% [3/25 partial responses (colorectal, cholangiocarcinoma, urothelial cancers), 90% confidence interval (CI) 3.4%-28.2%]. There was one additional partial response (urothelial cancer) in a patient with an unconfirmed ERBB2 copy number. Median PFS was 3.3 months (90% CI 2.0-4.1), and median OS 9.4 months (90% CI 5.0-18.9). Treatment-emergent adverse events were consistent with prior studies. There was no association between HER2 CN and response. CONCLUSIONS: HP was active in a selection of HER2-amplified tumors (non-breast/gastroesophageal) but did not meet the predefined efficacy benchmark. Additional strategies targeting HER2 and potential resistance pathways are warranted, especially in rare tumors.

Survival in patients with metastatic recurrent breast cancer after adjuvant chemotherapy: little evidence of improvement over the past 30 years.

BACKGROUND: Population-based studies have shown improved survival for patients diagnosed with metastatic breast cancer over time, presumably because of the availability of new and more effective therapies. The objective of the current study was to determine whether survival improved for patients who developed distant recurrence of breast cancer after receiving adjuvant therapy. METHODS: Adjuvant chemotherapy trials coordinated by the Eastern Cooperative Oncology Group that accrued patients between 1978 and 2002 were reviewed. Survival after distant disease recurrence was estimated for progressive time periods, and adjusted for baseline covariates in a Cox proportional hazards model. RESULTS: Of the 13,785 patients who received adjuvant chemotherapy in 11 trials, 3447 (25%) developed distant disease recurrence; the median survival after recurrence was 20 months (95% confidence interval, 19 months-21 months). Factors associated with inferior survival included a shorter distant recurrence-free interval (DRFI), estrogen receptor-negative and progesterone receptor-negative disease, the number of positive axillary lymph nodes present at the time of diagnosis, and black race (P < .0001 for all). When the time period of recurrence was added to the model, it was not found to be significantly associated with survival for the general population with disease recurrence. Survival improved over time only in those patients with hormone receptor-negative disease with a DRFI ≤ 3 years, both among the 5 most recent and the entire trial data sets (P = .01 and P = .05, respectively). CONCLUSIONS: In contrast to reports from population-based studies, no general improvement in survival was observed over the last 30 years for patients who developed distant disease recurrence after adjuvant chemotherapy after adjusting for DRFI. Improved survival for patients with hormone receptor-negative disease with a short DRFI suggests a benefit from trastuzumab.