Prediction of Effectiveness and Toxicities of Immune Checkpoint Inhibitors Using Real-World Patient Data.

PURPOSE: Although immune checkpoint inhibitors (ICIs) have improved outcomes in certain patients with cancer, they can also cause life-threatening immunotoxicities. Predicting immunotoxicity risks alongside response could provide a personalized risk-benefit profile, inform therapeutic decision making, and improve clinical trial cohort selection. We aimed to build a machine learning (ML) framework using routine electronic health record (EHR) data to predict hepatitis, colitis, pneumonitis, and 1-year overall survival. METHODS: Real-world EHR data of more than 2,200 patients treated with ICI through December 31, 2018, were used to develop predictive models. Using a prediction time point of ICI initiation, a 1-year prediction time window was applied to create binary labels for the four outcomes for each patient. Feature engineering involved aggregating laboratory measurements over appropriate time windows (60-365 days). Patients were randomly partitioned into training (80%) and test (20%) sets. Random forest classifiers were developed using a rigorous model development framework. RESULTS: The patient cohort had a median age of 63 years and was 61.8% male. Patients predominantly had melanoma (37.8%), lung cancer (27.3%), or genitourinary cancer (16.4%). They were treated with PD-1 (60.4%), PD-L1 (9.0%), and CTLA-4 (19.7%) ICIs. Our models demonstrate reasonably strong performance, with AUCs of 0.739, 0.729, 0.755, and 0.752 for the pneumonitis, hepatitis, colitis, and 1-year overall survival models, respectively. Each model relies on an outcome-specific feature set, though some features are shared among models. CONCLUSION: To our knowledge, this is the first ML solution that assesses individual ICI risk-benefit profiles based predominantly on routine structured EHR data. As such, use of our ML solution will not require additional data collection or documentation in the clinic.

Biomarker Inference and the Timing of Next-Generation Sequencing in a Multi-Institutional, Cross-Cancer Clinicogenomic Data Set.

PURPOSE: Observational clinicogenomic data sets, consisting of tumor next-generation sequencing (NGS) data linked to clinical records, are commonly used for cancer research. However, in real-world practice, oncologists frequently request NGS in search of treatment options for progressive cancer. The extent and impact of this dynamic on analysis of clinicogenomic research data are not well understood. METHODS: We analyzed clinicogenomic data for patients with non-small cell lung, colorectal, breast, prostate, pancreatic, or urothelial cancers in the American Association for Cancer Research Biopharmaceutical Consortium cohort. Associations between baseline and time-varying clinical characteristics and time from diagnosis to NGS were measured. To explore the impact of informative cohort entry on biomarker inference, statistical interactions between selected biomarkers and time to NGS with respect to overall survival were calculated. RESULTS: Among 7,182 patients, time from diagnosis to NGS varied significantly by clinical factors, including cancer type, calendar year of sequencing, institution, and age and stage at diagnosis. NGS rates also varied significantly by dynamic clinical status variables; in an adjusted model, compared with patients with stable disease at any given time after diagnosis, patients with progressive disease by imaging or oncologist assessment had higher NGS rates (hazard ratio for NGS, 1.61 [95% CI, 1.45 to 1.78] and 2.32 [95% CI, 2.01 to 2.67], respectively). Statistical interactions between selected biomarkers and time to NGS with respect to survival, potentially indicating biased biomarker inference results, were explored. CONCLUSION: To evaluate the appropriateness of a data set for a particular research question, it is crucial to measure associations between dynamic cancer status and the timing of NGS, as well as to evaluate interactions involving biomarkers of interest and NGS timing with respect to survival outcomes.

Natural History and Characteristics of ERBB2-mutated Hormone Receptor-positive Metastatic Breast Cancer: A Multi-institutional Retrospective Case-control Study from AACR Project GENIE.

PURPOSE: We wanted to determine the prognosis and the phenotypic characteristics of hormone receptor-positive advanced breast cancer tumors harboring an ERBB2 mutation in the absence of a HER2 amplification. EXPERIMENTAL DESIGN: We retrospectively collected information from the American Association of Cancer Research-Genomics Evidence Neoplasia Information Exchange registry database from patients with hormone receptor-positive, HER2-negative, ERBB2-mutated advanced breast cancer. Phenotypic and co-mutational features, as well as response to treatment and outcome were compared with matched control cases ERBB2 wild type. RESULTS: A total of 45 ERBB2-mutant cases were identified for 90 matched controls. The presence of an ERBB2 mutation was not associated with worse outcome determined by overall survival (OS) from first metastatic relapse. No significant differences were observed in phenotypic characteristics apart from higher lobular infiltrating subtype in the ERBB2-mutated group. ERBB2 mutation did not seem to have an impact in response to treatment or time-to-progression (TTP) to endocrine therapy compared with ERBB2 wild type. In the co-mutational analyses, CDH1 mutation was more frequent in the ERBB2-mutated group (FDR < 1). Although not significant, fewer co-occurring ESR1 mutations and more KRAS mutations were identified in the ERBB2-mutated group. CONCLUSIONS: ERBB2-activating mutation was not associated with a worse OS from time of first metastatic relapse, or differences in TTP on treatment as compared with a series of matched controls. Although not significant, differences in coexisting mutations (CDH1, ESR1, and KRAS) were noted between the ERBB2-mutated and the control group.

Learnings From Precision Clinical Trial Matching for Oncology Patients Who Received NGS Testing.

PURPOSE: Tumor next-generation sequencing reports typically generate trial recommendations for patients based on their diagnosis and genomic profile. However, these require additional refinement and prescreening, which can add to physician burden. We wanted to use human prescreening efforts to efficiently refine these trial options and also elucidate the high-value parameters that have a major impact on efficient trial matching. METHODS: Clinical trial recommendations were generated based on diagnosis and biomarker criteria using an informatics platform and were further refined by manual prescreening. The refined results were then compared with the initial trial recommendations and the reasons for false-positive matches were evaluated. RESULTS: Manual prescreening significantly reduced the number of false positives from the informatics generated trial recommendations, as expected. We found that trial-specific criteria, especially recruiting status for individual trial arms, were a high value parameter and led to the largest number of automated false-positive matches. CONCLUSION: Reflex clinical trial matching approaches that refine trial recommendations based on the clinical details as well as trial-specific criteria have the potential to help alleviate physician burden for selecting the most appropriate trial for their patient. Investing in publicly available resources that capture the recruiting status of a trial at the cohort or arm level would, therefore, allow us to make meaningful contributions to increase the clinical trial enrollments by eliminating false positives.

Characteristics and Outcome of AKT1 E17K-Mutant Breast Cancer Defined through AACR Project GENIE, a Clinicogenomic Registry.

AKT inhibitors have promising activity in AKT1 E17K-mutant estrogen receptor (ER)-positive metastatic breast cancer, but the natural history of this rare genomic subtype remains unknown. Utilizing AACR Project GENIE, an international clinicogenomic data-sharing consortium, we conducted a comparative analysis of clinical outcomes of patients with matched AKT1 E17K-mutant (n = 153) and AKT1-wild-type (n = 302) metastatic breast cancer. AKT1-mutant cases had similar adjusted overall survival (OS) compared with AKT1-wild-type controls (median OS, 24.1 vs. 29.9, respectively; P = 0.98). AKT1-mutant cases enjoyed longer durations on mTOR inhibitor therapy, an observation previously unrecognized in pivotal clinical trials due to the rarity of this alteration. Other baseline clinicopathologic features, as well as durations on other classes of therapy, were broadly similar. In summary, we demonstrate the feasibility of using a novel and publicly accessible clincogenomic registry to define outcomes in a rare genomically defined cancer subtype, an approach with broad applicability to precision oncology. SIGNIFICANCE: We delineate the natural history of a rare genomically distinct cancer, AKT1 E17K-mutant ER-positive breast cancer, using a publicly accessible registry of real-world patient data, thereby illustrating the potential to inform drug registration through synthetic control data.See related commentary by Castellanos and Baxi, p. 490.

American Association for Cancer Research Project Genomics Evidence Neoplasia Information Exchange: From Inception to First Data Release and Beyond-Lessons Learned and Member Institutions' Perspectives.

The American Association for Cancer Research (AACR) Project Genomics Evidence Neoplasia Information Exchange (GENIE) is an international data-sharing consortium focused on enabling advances in precision oncology through the gathering and sharing of tumor genetic sequencing data linked with clinical data. The project's history, operational structure, lessons learned, and institutional perspectives on participation in the data-sharing consortium are reviewed. Individuals involved with the inception and execution of AACR Project GENIE from each member institution described their experiences and lessons learned. The consortium was conceived in January 2014 and publicly released its first data set in January 2017, which consisted of 18,804 samples from 18,324 patients contributed by the eight founding institutions. Commitment and contributions from many individuals at AACR and the member institutions were crucial to the consortium's success. These individuals filled leadership, project management, informatics, data curation, contracts, ethics, and security roles. Many lessons were learned during the first 3 years of the consortium, including on how to gather, harmonize, and share data; how to make decisions and foster collaboration; and how to set the stage for continued participation and expansion of the consortium. We hope that the lessons shared here will assist new GENIE members as well as others who embark on the journey of forming a genomic data-sharing consortium.

Internet-Based Assessment of Oncology Health Care Professional Learning Style and Optimization of Materials for Web-Based Learning: Controlled Trial With Concealed Allocation.

BACKGROUND: Precision medicine has resulted in increasing complexity in the treatment of cancer. Web-based educational materials can help address the needs of oncology health care professionals seeking to understand up-to-date treatment strategies. OBJECTIVE: This study aimed to assess learning styles of oncology health care professionals and to determine whether learning style-tailored educational materials lead to enhanced learning. METHODS: In all, 21,465 oncology health care professionals were invited by email to participate in the fully automated, parallel group study. Enrollment and follow-up occurred between July 13 and September 7, 2015. Self-enrolled participants took a learning style survey and were assigned to the intervention or control arm using concealed alternating allocation. Participants in the intervention group viewed educational materials consistent with their preferences for learning (reading, listening, and/or watching); participants in the control group viewed educational materials typical of the My Cancer Genome website. Educational materials covered the topic of treatment of metastatic estrogen receptor-positive (ER+) breast cancer using cyclin-dependent kinases 4/6 (CDK4/6) inhibitors. Participant knowledge was assessed immediately before (pretest), immediately after (posttest), and 2 weeks after (follow-up test) review of the educational materials. Study statisticians were blinded to group assignment. RESULTS: A total of 751 participants enrolled in the study. Of these, 367 (48.9%) were allocated to the intervention arm and 384 (51.1%) were allocated to the control arm. Of those allocated to the intervention arm, 256 (69.8%) completed all assessments. Of those allocated to the control arm, 296 (77.1%) completed all assessments. An additional 12 participants were deemed ineligible and one withdrew. Of the 552 participants, 438 (79.3%) self-identified as multimodal learners. The intervention arm showed greater improvement in posttest score compared to the control group (0.4 points or 4.0% more improvement on average; P=.004) and a higher follow-up test score than the control group (0.3 points or 3.3% more improvement on average; P=.02). CONCLUSIONS: Although the study demonstrated more learning with learning style-tailored educational materials, the magnitude of increased learning and the largely multimodal learning styles preferred by the study participants lead us to conclude that future content-creation efforts should focus on multimodal educational materials rather than learning style-tailored content.

My Cancer Genome: Evaluating an Educational Model to Introduce Patients and Caregivers to Precision Medicine Information.

This study tested an innovative model for creating consumer-level content about precision medicine based on health literacy and learning style principles. "Knowledge pearl" videos, incorporating multiple learning modalities, were created to explain genetic and cancer medicine concepts. Cancer patients and caregivers (n=117) were randomized to view professional-level content directly from the My Cancer Genome (MCG) website (Group A; control), content from MCG with knowledge pearls embedded (Group B), or a consumer translation, targeted at the sixth grade level, with knowledge pearls embedded (Group C). A multivariate analysis showed that Group C, but not Group B, showed greater knowledge gains immediately after viewing the educational material than Group A. Statistically significant group differences in test performance were no longer observed three weeks later. These findings suggest that adherence to health literacy and learning style principles facilitates comprehension of precision medicine concepts and that ongoing review of the educational information is necessary.

Guiding Oncology Patients Through the Maze of Precision Medicine.

As the role of genomics in health care grows, patients increasingly require adequate genetic literacy to fully engage in their care. This study investigated a model for delivering consumer-friendly genetic information to improve understanding of precision medicine using health literacy and learning style principles. My Cancer Genome (MCG), a freely available cancer decision support tool, was used as a testbed. MCG content on a melanoma tumor mutation, BRAF V600E, was translated to a 6th-grade reading level, incorporating multiple learning modalities. A total of 90 patients and caregivers were recruited from a melanoma clinic at an academic medical center and randomized to 3 groups. Group A (control) received an exact copy of text from MCG. Group B was given the same content with hyperlinks to videos explaining key genetic concepts, identified and labeled by the team as knowledge pearls. Group C received the translated content with the knowledge pearls embedded. Changes in knowledge were measured through pre and post questionnaires. Group C showed the greatest improvement in knowledge. The study results demonstrate that providing information based on health literacy and learning style principles can improve patients' understanding of genetic concepts, thus increasing their likelihood of taking an active role in any decision making concerning their health.

The Path(way) Less Traveled: A Pathway-Oriented Approach to Providing Information about Precision Cancer Medicine on My Cancer Genome.

This perspective describes the motivation, development, and implementation of pathway-based content for My Cancer Genome, an online precision medicine knowledge resource describing clinical implications of genetic alterations in cancer. As researchers uncover more about cancer pathogenesis, we are learning more not only about the specific genes and proteins involved but also about how those genes and proteins interact with others along cell signaling pathways. This knowledge has led researchers and clinicians to begin to think about cancer therapy using a pathway-based approach. To facilitate this approach, My Cancer Genome used a list of more than 800 cancer-related genes to identify 20 cancer-relevant pathways and then created content focused on demonstrating the therapeutic relevance of these pathways.

Beyond histology: translating tumor genotypes into clinically effective targeted therapies.

Increased understanding of intertumoral heterogeneity at the genomic level has led to significant advancements in the treatment of solid tumors. Functional genomic alterations conferring sensitivity to targeted therapies can take many forms, and appropriate methods and tools are needed to detect these alterations. This review provides an update on genetic variability among solid tumors of similar histologic classification, using non-small cell lung cancer and melanoma as examples. We also discuss relevant technological platforms for discovery and diagnosis of clinically actionable variants and highlight the implications of specific genomic alterations for response to targeted therapy.

Identifying the status of genetic lesions in cancer clinical trial documents using machine learning.

BACKGROUND: Many cancer clinical trials now specify the particular status of a genetic lesion in a patient's tumor in the inclusion or exclusion criteria for trial enrollment. To facilitate search and identification of gene-associated clinical trials by potential participants and clinicians, it is important to develop automated methods to identify genetic information from narrative trial documents. METHODS: We developed a two-stage classification method to identify genes and genetic lesion statuses in clinical trial documents extracted from the National Cancer Institute's (NCI's) Physician Data Query (PDQ) cancer clinical trial database. The method consists of two steps: 1) to distinguish gene entities from non-gene entities such as English words; and 2) to determine whether and which genetic lesion status is associated with an identified gene entity. We developed and evaluated the performance of the method using a manually annotated data set containing 1,143 instances of the eight most frequently mentioned genes in cancer clinical trials. In addition, we applied the classifier to a real-world task of cancer trial annotation and evaluated its performance using a larger sample size (4,013 instances from 249 distinct human gene symbols detected from 250 trials). RESULTS: Our evaluation using a manually annotated data set showed that the two-stage classifier outperformed the single-stage classifier and achieved the best average accuracy of 83.7% for the eight most frequently mentioned genes when optimized feature sets were used. It also showed better generalizability when we applied the two-stage classifier trained on one set of genes to another independent gene. When a gene-neutral, two-stage classifier was applied to the real-world task of cancer trial annotation, it achieved a highest accuracy of 89.8%, demonstrating the feasibility of developing a gene-neutral classifier for this task. CONCLUSIONS: We presented a machine learning-based approach to detect gene entities and the genetic lesion statuses from clinical trial documents and demonstrated its use in cancer trial annotation. Such methods would be valuable for building information retrieval tools targeting gene-associated clinical trials.

Large-area spatially ordered arrays of gold nanoparticles directed by lithographically confined DNA origami.

The development of nanoscale electronic and photonic devices will require a combination of the high throughput of lithographic patterning and the high resolution and chemical precision afforded by self-assembly. However, the incorporation of nanomaterials with dimensions of less than 10 nm into functional devices has been hindered by the disparity between their size and the 100 nm feature sizes that can be routinely generated by lithography. Biomolecules offer a bridge between the two size regimes, with sub-10 nm dimensions, synthetic flexibility and a capability for self-recognition. Here, we report the directed assembly of 5-nm gold particles into large-area, spatially ordered, two-dimensional arrays through the site-selective deposition of mesoscopic DNA origami onto lithographically patterned substrates and the precise binding of gold nanocrystals to each DNA structure. We show organization with registry both within an individual DNA template and between components on neighbouring DNA origami, expanding the generality of this method towards many types of patterns and sizes.

Placement and orientation of individual DNA shapes on lithographically patterned surfaces.

Artificial DNA nanostructures show promise for the organization of functional materials to create nanoelectronic or nano-optical devices. DNA origami, in which a long single strand of DNA is folded into a shape using shorter 'staple strands', can display 6-nm-resolution patterns of binding sites, in principle allowing complex arrangements of carbon nanotubes, silicon nanowires, or quantum dots. However, DNA origami are synthesized in solution and uncontrolled deposition results in random arrangements; this makes it difficult to measure the properties of attached nanodevices or to integrate them with conventionally fabricated microcircuitry. Here we describe the use of electron-beam lithography and dry oxidative etching to create DNA origami-shaped binding sites on technologically useful materials, such as SiO(2) and diamond-like carbon. In buffer with approximately 100 mM MgCl(2), DNA origami bind with high selectivity and good orientation: 70-95% of sites have individual origami aligned with an angular dispersion (+/-1 s.d.) as low as +/-10 degrees (on diamond-like carbon) or +/-20 degrees (on SiO(2)).

Enzymatic ligation creates discrete multinanoparticle building blocks for self-assembly.

Enzymatic ligation of discrete nanoparticle-DNA conjugates creates nanoparticle dimer and trimer structures in which the nanoparticles are linked by single-stranded DNA, rather than by double-stranded DNA as in previous experiments. Ligation was verified by agarose gel and small-angle X-ray scattering. This capability was utilized in two ways: first, to create a new class of multiparticle building blocks for nanoscale self-assembly and, second, to develop a system that can amplify a population of discrete nanoparticle assemblies.

Discrete nanostructures of quantum dots/Au with DNA.

Nanostructures of colloidal CdSe/ZnS core/shell quantum dots (QDs) surrounded by a discrete number of Au nanoparticles were generated via DNA hybridization and purified by gel electrophoresis. Statistics from TEM analysis showed a high yield of designed structures. The distance between Au particles and QD, the number of Au around the central QD, and the size of Au and QD can be adjusted. Rationally designed structures such as these hold great promise for researching the physical interactions between semiconductor and Au nanoparticles and for developing more efficient nanoprobes.

Sorting fluorescent nanocrystals with DNA.

Semiconductor nanocrystals with narrow and tunable fluorescence are covalently linked to oligonucleotides. These biocompounds retain the properties of both nanocrystals and DNA. Therefore, different sequences of DNA can be coded with nanocrystals and still preserve their ability to hybridize to their complements. We report the case where four different sequences of DNA are linked to four nanocrystal samples having different colors of emission in the range of 530-640 nm. When the DNA-nanocrystal conjugates are mixed together, it is possible to sort each type of nanoparticle by using hybridization on a defined micrometer-size surface containing the complementary oligonucleotide. Detection of sorting requires only a single excitation source and an epifluorescence microscope. The possibility of directing fluorescent nanocrystals toward specific biological targets and detecting them, combined with their superior photostability compared to organic dyes, opens the way to improved biolabeling experiments, such as gene mapping on a nanometer scale or multicolor microarray analysis.