Summary of Research: Overall Survival with Osimertinib in Resected EGFR-Mutated NSCLC.

This is a summary of the original article ?Overall survival with osimertinib in resected EGFR-mutated NSCLC.Ë® Osimertinib blocks the activity of the epidermal growth factor receptor (EGFR) on cancer cells, causing cancer cell death and tumor shrinkage, and is an effective treatment for EGFR-mutated non-small cell lung cancer (NSCLC). The ADAURA study assessed the effects of osimertinib versus placebo in patients with EGFR-mutated (exon 19 deletion or L858R) early stage (IB-IIIA) NSCLC removed by surgery (resected). Previous results from ADAURA demonstrated that patients treated with osimertinib stayed alive and cancer-free (disease-free survival) significantly longer than patients who received placebo. Recent data showed the overall length of time patients were alive after starting treatment (overall survival). In both the primary stage II-IIIA and overall stage IB-IIIA populations, patients in the osimertinib group had a significant 51% reduction in the risk of death compared with the placebo group. The data demonstrated that osimertinib after surgery significantly improved overall survival in patients with resected, EGFR-mutated, stage IB-IIIA NSCLC.

Precision needle-punch tumor enrichment from paraffin blocks improves the detection of clinically actionable genomic alterations and biomarkers.

Background: While many molecular assays can detect mutations at low tumor purity and variant allele frequencies, complex biomarkers such as tumor mutational burden (TMB), microsatellite instability (MSI), and genomic loss of heterozygosity (gLOH) require higher tumor purity for accurate measurement. Scalable, quality-controlled, tissue-conserving methods to increase tumor nuclei percentage (TN%) from tumor specimens are needed for complex biomarkers and hence necessary to maximize patient matching to approved therapies or clinical trial enrollment. We evaluated the clinical utility and performance of precision needle-punch enrichment (NPE) compared with traditional razor blade macroenrichment of tumor specimens on molecular testing success. Methods: Pathologist-directed NPE was performed manually on formalin-fixed, paraffin embedded (FFPE) blocks. Quality control of target capture region and quantity of residual tumor in each tissue block was determined via a post-enrichment histologic slide recut. Resultant tumor purity and biomarker status were determined by the computational analysis pipeline component of the FDA-approved next-generation sequencing (NGS) assay, FoundationOne®CDx. Following NPE implementation for real-world clinical samples, assay performance and biomarker (MSI, TMB, gLOH) detection were analyzed. Results: In real-world clinical samples, enrichment rate via NPE was increased to ~50% over a 2.5-year period, exceeding the prior use of razor blade macro-enrichment (<30% of cases) prior to NPE implementation due to proven efficacy in generating high quality molecular results from marginal samples and the ease of use for both pathologist and histotechnologists. NPE was associated with lower test failures, higher computational tumor purity, and higher rates of successful TMB, MSI and gLOH determination when stratified by pre-enriched (incipient) tumor nuclei percentage. In addition, challenging cases in which tumor content was initially insufficient for testing were salvaged for analysis of biomarker status, gene amplification/deletion, and confident mutant or wild-type gene status determination. Conclusions: Pathologist-directed precision enrichment from tissue blocks (aka NPE) increases tumor purity, and consequently, yields a greater number of successful tests and complex biomarker determinations. Moreover, this process is rapid, safe, inexpensive, scalable, and conserves patient surgical pathology material. NPE may constitute best practice with respect to enriching tumor cells from low-purity specimens for biomarker detection in molecular laboratories.

Multi-omics Analysis Reveals Immune Features Associated with Immunotherapy Benefit in Patients with Squamous Cell Lung Cancer from Phase III Lung-MAP S1400I Trial.

PURPOSE: Identifying molecular and immune features to guide immune checkpoint inhibitor (ICI)-based regimens remains an unmet clinical need. EXPERIMENTAL DESIGN: Tissue and longitudinal blood specimens from phase III trial S1400I in patients with metastatic squamous non-small cell carcinoma (SqNSCLC) treated with nivolumab monotherapy (nivo) or nivolumab plus ipilimumab (nivo+ipi) were subjected to multi-omics analyses including multiplex immunofluorescence (mIF), nCounter PanCancer Immune Profiling Panel, whole-exome sequencing, and Olink. RESULTS: Higher immune scores from immune gene expression profiling or immune cell infiltration by mIF were associated with response to ICIs and improved survival, except regulatory T cells, which were associated with worse overall survival (OS) for patients receiving nivo+ipi. Immune cell density and closer proximity of CD8+GZB+ T cells to malignant cells were associated with superior progression-free survival and OS. The cold immune landscape of NSCLC was associated with a higher level of chromosomal copy-number variation (CNV) burden. Patients with LRP1B-mutant tumors had a shorter survival than patients with LRP1B-wild-type tumors. Olink assays revealed soluble proteins such as LAMP3 increased in responders while IL6 and CXCL13 increased in nonresponders. Upregulation of serum CXCL13, MMP12, CSF-1, and IL8 were associated with worse survival before radiologic progression. CONCLUSIONS: The frequency, distribution, and clustering of immune cells relative to malignant ones can impact ICI efficacy in patients with SqNSCLC. High CNV burden may contribute to the cold immune microenvironment. Soluble inflammation/immune-related proteins in the blood have the potential to monitor therapeutic benefit from ICI treatment in patients with SqNSCLC.

Safety profile of pembrolizumab monotherapy based on an aggregate safety evaluation of 8937 patients.

BACKGROUND: Pembrolizumab has a manageable safety profile as described in its label, which was primarily based on 2799 patients who participated in clinical trials for melanoma or non-small cell lung cancer. Here, we evaluated the safety of pembrolizumab in a broader population of patients from 31 advanced cancer clinical trials across 19 cancer types. METHODS: Safety was analyzed in patients who received at least one dose of pembrolizumab (200 mg every 3 weeks [Q3W], 10 mg/kg Q2W or Q3W, or 2 mg/kg Q3W). Adverse events (AEs) and immune-mediated AEs and infusion reactions were evaluated. RESULTS: Safety data from 8937 patients in 31 trials of pembrolizumab monotherapy were pooled (median, seven administrations; range, 1-59). Median duration on treatment was 4.1 months (range, 0.03-40.1). AEs occurred in 96.6% of patients. Grade 3-5 AEs occurred in 50.6% of patients. AEs led to pembrolizumab discontinuation in 12.7% of patients and death in 5.9%. Immune-mediated AEs and infusion reactions occurred in 23.7% of patients (4.6% experienced multiple immune-mediated AEs/infusion reactions) and led to pembrolizumab discontinuation in 3.6% and death in 0.2%. Grade 3-5 immune-mediated AEs occurred in 6.3% of patients. Serious immune-mediated AEs and infusion reactions occurred in 6.0% of patients. Median time to immune-mediated AE onset was 85 days (range, 13-163). Of 2657 immune-mediated AEs, 22.3% were initially treated with prednisone ≥ 40 mg/day or equivalent, and 8.3% were initially treated with lower steroid doses. CONCLUSIONS: This pooled analysis of 31 clinical trials showed that pembrolizumab has a consistent safety profile across indications.

Understanding health-related quality of life measures used in early-stage non-small cell lung cancer clinical trials: A review.

Health-related quality of life (HRQoL) is an important consideration in cancer clinical research, which can be substantially influenced by cancer treatment procedures and medications. The treatment landscape for early-stage (stage I-III) non-small cell lung cancer (NSCLC) is rapidly evolving. In this light, it is important to evaluate the most suitable instruments for HRQoL assessment and timing. Given there is often a requirement for patients with early-stage disease to receive long-term treatment to reduce the risk of disease recurrence after surgery, maintenance or improvement in HRQoL is an important goal of both neoadjuvant and adjuvant treatments. Key challenges with assessing HRQoL relate to the suitability of existing instruments to measure relevant treatment-related adverse effects, consistency in HRQoL assessment approach between similar studies, gaps in data collection and reporting, and interpretation of longitudinal data. Frequent assessments during and after treatment are warranted to capture the true impact of treatment and disease progression on HRQoL, and changes in the relative importance of these factors over time. There is scope for improving existing HRQoL approaches, including ease of use and integration of digital tools to facilitate analysis and interpretation, to enhance the experience of both patients and healthcare professionals. In this narrative review, we discuss key considerations for HRQoL assessment and evaluate the tools currently available to measure HRQoL in NSCLC, many of which were designed with advanced disease in mind. We focus on the key challenges of measuring HRQoL for the specific needs of patients with early-stage disease, and consider future perspectives, to determine the most appropriate HRQoL instruments and analysis methods to use in early-stage NSCLC clinical trials.1.

Effect of a Personalized Tobacco Treatment Intervention on Smoking Abstinence in Individuals Eligible for Lung Cancer Screening.

INTRODUCTION: To determine whether personalized gain-framed messaging and biomarker feedback related to tobacco cessation or reduction decrease smoking behavior in patients undergoing or eligible for lung cancer screening. METHODS: Between 2016 and 2020, 188 patients were enrolled in a two-phase, sequential, randomized controlled trial. Phase 1 evaluated whether standard of care (SC) (five in-person counseling sessions and 8 weeks of nicotine patch) plus gain-framed messaging (GFM) versus SC would increase 8-week biochemically verified smoking cessation rates. In 143 participants randomized in phase 2, we tested whether feedback on smoking-related biomarkers would reduce 6-month self-reported number of cigarettes smoked per day compared with a no feedback control. Chi-square test and mixed effects repeated measures analyses were used to evaluate group differences. RESULTS: Participants were 62.5 ± 5.6 (mean ± SD) years of age, had a 50.3 ± 21 pack-year smoking history, and were smoking 16.9 ± 9.9 cigarettes per day. At 8 weeks, there was no difference in quit rates between those randomized to SC plus GFM (n = 15 of 93, 16.1%) and those randomized to SC (n = 16 of 95, 16.8%), with p equals to 0.90. At the 6-month post-randomization follow-up, number of cigarettes smoked per day was similar in the feedback (least-squares mean = 7.5, 95% confidence interval: 6.0-9.1) and no feedback arms (7.7, 95% confidence interval: 6.2-9.3), with p equals to 0.87. CONCLUSIONS: Gain-framed messaging and health feedback did not significantly improve quit rates relative to comprehensive standard of care. Nevertheless, the overall program achieved clinically meaningful smoking quit rates in this older high pack-year cohort, highlighting the importance of intensive tobacco treatment for patients undergoing lung cancer screening. CLINICAL TRIAL REGISTERED WITH CLINICALTRIALS.GOV: NCT02658032.

Novel Approach to Accelerate Lung Cancer Research: Lung-MAP and the Potential of Public-Private Partnerships.

The National Cancer Institute recently found that death rates for non-small cell lung cancer (NSCLC) have been reduced by over 6% overall in recent years. This reduction in mortality has been accompanied by an average increase in overall survival and largely credited to the therapeutic advancements for the effective treatment of NSCLC. Numerous molecular alterations have been identified in NSCLC that have enabled the development of new drugs capable of targeting these changes and efficiently kill cancerous cells. New treatments to modulate patients' immune systems have been shown to be effective in stimulating natural immune cells to have an improved anti-cancer effect. While these types of approaches to treat cancer are providing new options for patients, leadership from the Food and Drug Administration (FDA) recognized that the expansion of targeted therapy in NSCLC presented significant promise, but evaluation of the safety and efficacy of these new drugs would be slowed if new models for conducting clinical studies were not identified. Specifically, the FDA recommended that a comprehensive approach be implemented to identify the patients that are the best candidates for these, and other new treatments based upon the molecular characteristics of their tumors, and more efficiently conduct the clinical studies necessary to evaluate the safety and efficacy of new drugs. To address this growing challenge, leading lung cancer experts and stakeholders across academia, government, industry, and patient advocacy came together to design a clinical research approach that could serve as a sustainable infrastructure for new lung cancer treatments called the Lung Cancer Master Protocol.

Representativeness of Patients Enrolled in the Lung Cancer Master Protocol (Lung-MAP).

PURPOSE: Lung Cancer Master Protocol (Lung-MAP), a public-private partnership, established infrastructure for conducting a biomarker-driven master protocol in molecularly targeted therapies. We compared characteristics of patients enrolled in Lung-MAP with those of patients in advanced non-small-cell lung cancer (NSCLC) trials to examine if master protocols improve trial access. METHODS: We examined patients enrolled in Lung-MAP (2014-2020) according to sociodemographic characteristics. Proportions for characteristics were compared with those for a set of advanced NSCLC trials (2001-2020) and the US advanced NSCLC population using SEER registry data (2014-2018). Characteristics of patients enrolled in Lung-MAP treatment substudies were examined in subgroup analysis. Two-sided tests of proportions at an alpha of .01 were used for all comparisons. RESULTS: A total of 3,556 patients enrolled in Lung-MAP were compared with 2,215 patients enrolled in other NSCLC studies. Patients enrolled in Lung-MAP were more likely to be 65 years and older (57.2% v 46.3%; P < .0001), from rural areas (17.3% v 14.4%; P = .004), and from socioeconomically deprived neighborhoods (42.2% v 36.7%, P < .0001), but less likely to be female (38.6% v 47.2%; P < .0001), Asian (2.8% v 5.1%; P < .0001), or Hispanic (2.4% v 3.8%; P = .003). Among patients younger than 65 years, Lung-MAP enrolled more patients using Medicaid/no insurance (27.6% v 17.8%; P < .0001). Compared with the US advanced NSCLC population, Lung-MAP under represented patients 65 years and older (57.2% v 69.8%; P < .0001), females (38.6% v 46.0%; P < .0001), and racial or ethnic minorities (14.8% v 21.5%; P < .0001). CONCLUSION: Master protocols may improve access to trials using novel therapeutics for older patients and socioeconomically vulnerable patients compared with conventional trials, but specific patient exclusion criteria influenced demographic composition. Further research examining participation barriers for under represented racial or ethnic minorities in precision medicine clinical trials is warranted.

Biomarker-directed, pembrolizumab-based combination therapy in non-small cell lung cancer: phase 2 KEYNOTE-495/KeyImPaCT trial interim results.

Although pembrolizumab confers clinical benefit in non-small cell lung cancer (NSCLC), only a subset of patients will respond due to a heterogenous tumor microenvironment. KEYNOTE-495/KeyImPaCT is an ongoing biomarker-directed, adaptively randomized phase 2 study investigating first-line pembrolizumab (200 mg every 3 weeks) + lenvatinib (20 mg daily), anti-CTLA-4 quavonlimab (25 mg every 6 weeks) or anti-LAG-3 favezelimab (200 mg or 800 mg every 3 weeks) in advanced NSCLC. Patients were categorized by T-cell-inflamed gene expression profile (TcellinfGEP) and tumor mutational burden (TMB) status and randomly assigned 1:1:1 to receive pembrolizumab + lenvatinib, pembrolizumab + quavonlimab or pembrolizumab + favezelimab. The primary outcome was investigator-assessed objective response rate (ORR) per Response Evaluation Criteria in Solid Tumors version 1.1 using pre-specified efficacy thresholds for each biomarker-defined subgroup (>5% (TcellinfGEPlowTMBnon-high (group I)), >20% (TcellinfGEPlowTMBhigh (group II) and TcellinfGEPnon-lowTMBnon-high (group III)) and >45% (TcellinfGEPnon-lowTMBhigh (group IV))). Secondary outcomes were progression-free survival, overall survival and safety. At data cutoff, ORR ranges were 0-12.0% in group I, 27.3-33.3% in group II, 13.6-40.9% in group III and 50.0-60.0% in group IV. ORR with pembrolizumab + lenvatinib in group III met the pre-specified efficacy threshold. The safety profile of each treatment arm was consistent with the known safety profile of each combination. These data demonstrate the feasibility of prospective TcellinfGEP and TMB assessment to study the clinical activity of first-line pembrolizumab-based combination therapies in advanced NSCLC. ClinicalTrials.gov registration: NCT03516981 .

Overall Survival with Osimertinib in Resected EGFR-Mutated NSCLC.

BACKGROUND: Among patients with resected, epidermal growth factor receptor (EGFR)-mutated, stage IB to IIIA non-small-cell lung cancer (NSCLC), adjuvant osimertinib therapy, with or without previous adjuvant chemotherapy, resulted in significantly longer disease-free survival than placebo in the ADAURA trial. We report the results of the planned final analysis of overall survival. METHODS: In this phase 3, double-blind trial, we randomly assigned eligible patients in a 1:1 ratio to receive osimertinib (80 mg once daily) or placebo until disease recurrence was observed, the trial regimen was completed (3 years), or a discontinuation criterion was met. The primary end point was investigator-assessed disease-free survival among patients with stage II to IIIA disease. Secondary end points included disease-free survival among patients with stage IB to IIIA disease, overall survival, and safety. RESULTS: Of 682 patients who underwent randomization, 339 received osimertinib and 343 received placebo. Among patients with stage II to IIIA disease, the 5-year overall survival was 85% in the osimertinib group and 73% in the placebo group (overall hazard ratio for death, 0.49; 95.03% confidence interval [CI], 0.33 to 0.73; P<0.001). In the overall population (patients with stage IB to IIIA disease), the 5-year overall survival was 88% in the osimertinib group and 78% in the placebo group (overall hazard ratio for death, 0.49; 95.03% CI, 0.34 to 0.70; P<0.001). One new serious adverse event, pneumonia related to coronavirus disease 2019, was reported after the previously published data-cutoff date (the event was not considered by the investigator to be related to the trial regimen, and the patient fully recovered). Adjuvant osimertinib had a safety profile consistent with that in the primary analysis. CONCLUSIONS: Adjuvant osimertinib provided a significant overall survival benefit among patients with completely resected, EGFR-mutated, stage IB to IIIA NSCLC. (Funded by AstraZeneca; ADAURA ClinicalTrials.gov number, NCT02511106.).

Three-Year Safety, Tolerability, and Health-Related Quality of Life Outcomes of Adjuvant Osimertinib in Patients With Resected Stage IB to IIIA EGFR-Mutated NSCLC: Updated Analysis From the Phase 3 ADAURA Trial.

INTRODUCTION: In ADAURA, adjuvant osimertinib significantly improved disease-free survival versus placebo in resected stage IB to IIIA EGFR-mutated NSCLC. We report in-depth analyses of three-year safety, tolerability, and health-related quality of life (HRQoL) from ADAURA. METHODS: Patients were randomized 1:1 to osimertinib 80 mg or placebo once daily for up to 3 years. Safety assessments were performed at baseline, week 2, week 4, week 12, and every 12 weeks until treatment completion or discontinuation, and 28 days after treatment was stopped. The SF-36 survey measured HRQoL at baseline, week 12, week 24, and every 24 weeks until recurrence, treatment completion or discontinuation. Data cutoff: April 11, 2022. RESULTS: Safety and HRQoL analysis sets: osimertinib, n = 337 and n = 339; placebo, n = 343 each. Median (range) total exposure duration was longer with osimertinib versus placebo: 35.8 (0-38) versus 25.1 (0-39) months. Most adverse events (AEs) were first reported within 12 months of starting treatment (osimertinib 97%, placebo 86%). AEs leading to dose reduction, interruption or discontinuation were reported in 12%, 27% and 13% respectively of patients with osimertinib; 1%, 13% and 3% with placebo. Stomatitis and diarrhea were the most common AEs leading to osimertinib dose reduction or interruption; interstitial lung disease was the most common leading to osimertinib discontinuation (per protocol). There were no differences in time to deterioration for SF-36 physical, mental component summaries between osimertinib and placebo. CONCLUSIONS: No new safety signals were reported and HRQoL was maintained with 3 years of adjuvant osimertinib treatment. Combined with significant efficacy benefit, these data further support adjuvant osimertinib in stage IB to IIIA EGFR-mutated NSCLC.

Randomized Phase II Trial Comparing Bevacizumab Plus Carboplatin and Paclitaxel With Carboplatin and Paclitaxel Alone in Previously Untreated Locally Advanced or Metastatic Non-Small-Cell Lung Cancer.

PURPOSE: To investigate the efficacy and safety of bevacizumab plus carboplatin and paclitaxel in patients with advanced or recurrent non-small-cell lung cancer. PATIENTS AND METHODS: In a phase II trial, 99 patients were randomly assigned to bevacizumab 7.5 (n = 32) or 15 mg/kg (n = 35) plus carboplatin (area under the curve = 6) and paclitaxel (200 mg/m2) every 3 weeks or carboplatin and paclitaxel alone (n = 32). Primary efficacy end points were time to disease progression and best confirmed response rate. On disease progression, patients in the control arm had the option to receive single-agent bevacizumab 15 mg/kg every 3 weeks. RESULTS: Compared with the control arm, treatment with carboplatin and paclitaxel plus bevacizumab (15 mg/kg) resulted in a higher response rate (31.5% v 18.8%), longer median time to progression (7.4 v 4.2 months) and a modest increase in survival (17.7 v 14.9 months). Of the 19 control patients that crossed over to single-agent bevacizumab, five experienced stable disease, and 1-year survival was 47%. Bleeding was the most prominent adverse event and was manifested in two distinct clinical patterns; minor mucocutaneous hemorrhage and major hemoptysis. Major hemoptysis was associated with squamous cell histology, tumor necrosis and cavitation, and disease location close to major blood vessels. CONCLUSION: Bevacizumab in combination with carboplatin and paclitaxel improved overall response and time to progression in patients with advanced or recurrent non-small-cell lung cancer. Patients with nonsquamous cell histology appear to be a subpopulation with improved outcome and acceptable safety risks.

Novel Approaches for Dynamic Visualization of Adverse Event Data in Oncology Clinical Trials: A Case Study Using Immunotherapy Trial S1400-I (SWOG).

PURPOSE: Clinical trial adverse event (AE) data are increasingly complex and high-dimensional, especially for trials evaluating novel targeted agents and immunotherapies. Standard approaches to summarize and analyze AEs remain generally tabular, failing to describe the nature of AEs. Novel dynamic and data visualization methods are needed to enable a more comprehensive assessment of the overall toxicity profile of treatments. METHODS: We developed methods for visualizing the numerous categorizations and types of AEs along with a dynamic approach to better reflect its highly dimensional nature without sacrificing the reporting of rare events. Circular plots displaying the proportion of maximal-grade AEs by system organ classes (SOCs) and butterfly plots displaying the proportion of AEs by severity for each AE term were developed to enable comparisons of AE patterns by treatment arm. These approaches were applied to a randomized phase III trial (S1400I; ClinicalTrials.gov identifier: NCT02785952) comparing nivolumab with nivolumab plus ipilimumab in patients with stage IV squamous non-small-cell lung cancer. RESULTS: Our visualizations revealed that patients randomly assigned to nivolumab and ipilimumab had higher rates of grade 3 or higher AEs compared with nivolumab alone for several SOCs, including musculoskeletal (5.6% v 0.8%), skin (5.6% v 0.8%), vascular (5.6% v 1.6%), and cardiac (4% v 1.6%) toxicities. They also suggested a pattern of higher prevalence of moderate GI and endocrine toxicities and showed that although the rates of cardiac and neurologic toxicities were similar, the types of events were discordant. CONCLUSION: The graphical approaches we proposed enable a more comprehensive and intuitive evaluation of toxicity types by treatment groups, which is not apparent in tabular and descriptive reporting methods.

Genomic and transcriptomic analysis of checkpoint blockade response in advanced non-small cell lung cancer.

Anti-PD-1/PD-L1 agents have transformed the treatment landscape of advanced non-small cell lung cancer (NSCLC). To expand our understanding of the molecular features underlying response to checkpoint inhibitors in NSCLC, we describe here the first joint analysis of the Stand Up To Cancer-Mark Foundation cohort, a resource of whole exome and/or RNA sequencing from 393 patients with NSCLC treated with anti-PD-(L)1 therapy, along with matched clinical response annotation. We identify a number of associations between molecular features and outcome, including (1) favorable (for example, ATM altered) and unfavorable (for example, TERT amplified) genomic subgroups, (2) a prominent association between expression of inducible components of the immunoproteasome and response and (3) a dedifferentiated tumor-intrinsic subtype with enhanced response to checkpoint blockade. Taken together, results from this cohort demonstrate the complexity of biological determinants underlying immunotherapy outcomes and reinforce the discovery potential of integrative analysis within large, well-curated, cancer-specific cohorts.

Comparing Detection Schemes for Adversarial Images against Deep Learning Models for Cancer Imaging.

Deep learning (DL) models have demonstrated state-of-the-art performance in the classification of diagnostic imaging in oncology. However, DL models for medical images can be compromised by adversarial images, where pixel values of input images are manipulated to deceive the DL model. To address this limitation, our study investigates the detectability of adversarial images in oncology using multiple detection schemes. Experiments were conducted on thoracic computed tomography (CT) scans, mammography, and brain magnetic resonance imaging (MRI). For each dataset we trained a convolutional neural network to classify the presence or absence of malignancy. We trained five DL and machine learning (ML)-based detection models and tested their performance in detecting adversarial images. Adversarial images generated using projected gradient descent (PGD) with a perturbation size of 0.004 were detected by the ResNet detection model with an accuracy of 100% for CT, 100% for mammogram, and 90.0% for MRI. Overall, adversarial images were detected with high accuracy in settings where adversarial perturbation was above set thresholds. Adversarial detection should be considered alongside adversarial training as a defense technique to protect DL models for cancer imaging classification from the threat of adversarial images.

Quality-of-life outcomes and risk prediction for patients randomized to nivolumab plus ipilimumab vs nivolumab on LungMAP-S1400I.

BACKGROUND: An important issue for patients with cancer treated with novel therapeutics is how they weigh the effects of treatment on survival and quality of life (QOL). We compared QOL in patients enrolled to SWOG S1400I, a substudy of the LungMAP biomarker-driven master protocol. METHODS: SWOG S1400I was a randomized phase III trial comparing nivolumab plus ipilimumab vs nivolumab for treatment of immunotherapy-naïve disease in advanced squamous cell lung cancer. The primary endpoint was the MD Anderson Symptom Inventory-Lung Cancer severity score at week 7 and week 13 with a target difference of 1.0 points, assessed using multivariable linear regression. A composite risk model for progression-free and overall survival was derived using best-subset selection. RESULTS: Among 158 evaluable patients, median age was 67.6 years and most were male (66.5%). The adjusted MD Anderson Symptom Inventory-Lung Cancer severity score was 0.04 points (95% confidence interval [CI] = -0.44 to 0.51 points; P = .89) at week 7 and 0.12 points (95% CI = -0.41 to 0.65; P = .66) at week 13. A composite risk model showed that patients with high levels of appetite loss and shortness of breath had a threefold increased risk of progression or death (hazard ratio [HR] = 3.06, 95% CI = 1.88 to 4.98; P < .001) and that those with high levels of both appetite loss and work limitations had a fivefold increased risk of death (HR = 5.60, 95% CI = 3.27 to 9.57; P < .001)-compared with those with neither risk category. CONCLUSIONS: We found no evidence of a benefit of ipilimumab added to nivolumab compared with nivolumab alone for QOL in S1400I. A risk model identified patients at high risk of poor survival, demonstrating the prognostic relevance of baseline patient-reported outcomes even in those with previously treated advanced cancer.