Highly variable timing renders immunotherapy efficacy and toxicity impractical biomarkers of one another in clinical practice.

Background: A useful clinical biomarker requires not only association but also a consistent temporal relationship. For instance, chemotherapy-induced neutropenia and epidermal growth-factor inhibitor-related acneiform rash both occur within weeks of treatment initiation, thereby providing information prior to efficacy assessment. Although immune checkpoint inhibitor (ICI)-associated immune-related adverse events (irAE) have been associated with therapeutic benefit, irAE may have delayed and highly variable onset. To determine whether ICI efficacy and irAE could serve as clinically useful biomarkers for predicting each other, we determined the temporal relationship between initial efficacy assessment and irAE onset in a diverse population treated with ICI. Methods: Using two-sided Fisher exact and Cochran-Armitage tests, we determined the relative timing of initial efficacy assessment and irAE occurrence in a cohort of 155 ICI-treated patients (median age 68 years, 40% women). Results: Initial efficacy assessment was performed a median of 50 days [interquartile range (IQR) 39-59 days] after ICI initiation; median time to any irAE was 77 days (IQR 28-145 days) after ICI initiation. Median time to first irAE was 42 days (IQR 20-88 days). Overall, 58% of any irAE and 47% of first irAE occurred after initial efficacy assessment. For clinically significant (grade ≥2) irAE, 60% of any and 53% of first occurred after initial efficacy assessment. The likelihood of any future irAE did not differ according to response (45% for complete or partial response vs. 47% for other cases; P=1). In landmark analyses controlling for clinical and toxicity follow-up, patients demonstrating greater tumor shrinkage at initial efficacy assessment were more likely to develop future grade ≥2 (P=0.05) and multi-organ (P=0.02) irAE. Conclusions: In contrast to that seen with chemotherapy and molecularly targeted therapies, the temporal relationship between ICI efficacy and toxicity is complex and bidirectional. In practice, neither parameter can be routinely relied on as a clinical biomarker to predict the other.

Longitudinal Experience With and Impressions of COVID-19-Related Clinical Research Changes.

PURPOSE: The COVID-19 pandemic has led to profound changes in clinical research, including remote consent, telehealth, off-site procedures, shipment of therapy, and remote study monitoring. We assessed longitudinal perceptions of these adjustments among clinical research professionals. METHODS: We distributed an anonymous survey assessing experiences, perceptions, and recommendations regarding COVID-19-related clinical research adjustments to cancer clinical research office personnel in May 2020 and again in November 2020. Responses were compared using Fisher's exact and Mann-Whitney U tests. RESULTS: A total of 90 of 102 invited research personnel (88%) responded. Fifty-three (59%) reported participating in both initial and follow-up surveys. The proportion of respondents reporting personal experience with COVID-19-related adjustments increased over time, particularly for remote initial consent (29% v 4%), remote reconsent (24% v 9%), and remote study monitoring (36% v 22%). Perceived impact of COVID-19-related adjustments on data quality (P = .02) and patient experience (P = .002) improved significantly. However, perceived effect on patient safety (P = .02) and respondent's experience (P = .09) became less favorable. Individuals with personal experience with the adjustment were more likely to recommend continuing remote consent (62% v 38%; P = .04), remote monitoring (69% v 45%; P = .05), and therapy shipment (67% v 35%; P = .01) after the COVID-19 pandemic, with nonsignificant trends for off-site diagnostics (44% v 24%; P = .13) and telehealth visits (66% v 45%; P = .08). CONCLUSION: More than 6 months into the global pandemic, perceptions of COVID-19-related clinical research changes remain favorable. Experienced individuals are more likely to recommend that these changes continue in the future.

Experience, Perceptions, and Recommendations Concerning COVID-19-Related Clinical Research Adjustments.

BACKGROUND: During the COVID-19 public health emergency, the FDA and NIH altered clinical trial requirements to protect participants and manage study conduct. Given their detailed knowledge of research protocols and regular contact with patients, clinicians, and sponsors, clinical research professionals offer important perspectives on these changes. METHODS: We developed and distributed an anonymous survey assessing COVID-19-related clinical trial adjustment experiences, perceptions, and recommendations to Clinical Research Office personnel at the Harold C. Simmons Comprehensive Cancer Center. Responses were compared using the Fisher exact test. RESULTS: A total of 94 of 109 contacted research personnel (87%) responded. Among these individuals, 58% had >5 years' professional experience in clinical research, and 56% had personal experience with a COVID-19-related change. Respondents perceived that these changes had a positive impact on patient safety; treatment efficacy; patient and staff experience; and communication with patients, investigators, and sponsors. More than 90% felt that positive changes should be continued after COVID-19. For remote consent, telehealth, therapy shipment, off-site diagnostics, and remote monitoring, individuals with personal experience with the specific change and individuals with >5 years' professional experience were numerically more likely to recommend continuing the adjustment, and these differences were significant for telehealth (P=.04) and therapy shipment (P=.02). CONCLUSIONS: Clinical research professionals perceive that COVID-19-related clinical trial adjustments positively impact multiple aspects of study conduct. Those with greatest experience-both specific to COVID-19-related changes and more generally-are more likely to recommend that these adjustments continue in the future.

Ensemble QSAR Modeling to Predict Multispecies Fish Toxicity Lethal Concentrations and Points of Departure.

QSAR modeling can be used to aid testing prioritization of the thousands of chemical substances for which no ecological toxicity data are available. We drew on the U.S. Environmental Protection Agency's ECOTOX database with additional data from ECHA to build a large data set containing in vivo test data on fish for thousands of chemical substances. This was used to create QSAR models to predict two types of end points: acute LC50 (median lethal concentration) and points of departure similar to the NOEC (no observed effect concentration) for any duration (named the "LC50" and "NOEC" models, respectively). These models used study covariates, such as species and exposure route, as features to facilitate the simultaneous use of varied data types. A novel method of substituting taxonomy groups for species dummy variables was introduced to maximize generalizability to different species. A stacked ensemble of three machine learning methods-random forest, gradient boosted trees, and support vector regression-was implemented to best make use of a large data set with many descriptors. The LC50 and NOEC models predicted end points within 1 order of magnitude 81% and 76% of the time, respectively, and had RMSEs of roughly 0.83 and 0.98 log10(mg/L), respectively. Benchmarks against the existing TEST and ECOSAR tools suggest improved prediction accuracy.

Ensemble dynamics and the emergence of correlations in one- and two-dimensional wave turbulence.

We investigate statistical properties of wave turbulence by monitoring the dynamics of ensembles of trajectories. The system under investigation is a simplified model for surface gravity waves in one and two dimensions with a square-root dispersion and a four-wave interaction term. The simulations of decaying turbulence confirm the Kolmogorov-Zakharov spectral power distribution of wave turbulence theory. Fourth-order correlations are computed numerically as ensemble averages of trajectories. The shape, scaling, and time evolution of the correlations agree with the predictions of wave turbulence theory.

Transition of weak wave turbulence to wave turbulence with intermittent collapses.

We study the dynamics of one-dimensional nonlinear waves with a square-root dispersion. This dispersion allows strong interactions of distant modes in wave-number space, and it leads to a modulational instability of a carrier wave interacting with distant sidebands. Weak wave turbulence is found when the system is damped and weakly driven. A driving force that exceeds a critical strength leads to wave collapses coexisting with weak wave turbulence. We explain this transition behavior with the modulational instability of waves with the highest power: Below the threshold the instability is suppressed by the external long-wave damping force. Above the threshold the instability initiates wave collapses.