Indirect modeling of derived outcomes: Are minor prediction discrepancies a cause for concern?

It is often a goal of model development to predict data from which a variety of outcomes can be derived, such as threshold-based categorization or change from baseline (CFB) transformations. This approach can improve power or support multiple decisions. Because these derivations are indirectly predicted from the model, they are valuable tests for misspecification when used in visual or numeric predictive checks (V/NPCs). However, derived outcome V/NPCs (especially if primary or key secondary) are often overly scrutinized and held to an uncommon standard when comparing model predictions to point estimates, even if by conventional standards both the directly and indirectly modeled data are captured well. Here, simulations of directly modeled data were used to determine where apparent issues in V/NPCs of derived outcomes are expected. Two types of datasets were simulated: (1) a simple pre-post study and (2) pharmacokinetic/pharmacodynamic data from a dose-ranging study. A psoriasis exposure-response model case study was also assessed. V/NPCs were generated on the raw data, CFB data, and placebo-corrected CFB (dCFB) data, and binned summary statistics of the observed data for each trial were graded as being strongly or weakly supportive of a predictive model (within the interquartile range or the 95% central distribution of all simulated trials, respectively). In all cases, the strength of support in direct data V/NPCs was minimally related to that in derived outcome V/NPCs. There are myriad benefits to modeling the underlying data of a derived measure, and these results support caution in discarding adequate models based on overly strict derived measure predictive checks.

Simulating Healthy Participant Pharmacokinetics for Renal and Hepatic Impairment Studies: Retrospective Assessment of the Approach.

The recruitment of a parallel, healthy participants (HPs) arm in renal and hepatic impairment (RI and HI) studies is a common strategy to assess differences in pharmacokinetics. Limitations in this approach include the underpowered estimate of exposure differences and the use of the drug in a population for which there is no benefit. Recently, a method was published by Purohit et. al. (2023) that leveraged prior population pharmacokinetic (PopPK) modeling-based simulation to infer the distribution of exposure ratios between the RI/HI arms and HPs. The approach was successful, but it was a single example with a robust model having several iterations of development and fitting to extensive HP data. To test in more studies and models at different stages of development, our catalogue of RI/HI studies was searched, and those with suitable properties and from programs with available models were analyzed with the simulation approach. There were 9 studies included in the analysis. Most studies were associated with models that would have been available at the time (ATT) of the study, and all had a current, final model. For 3 studies, the HP PK was not predicted well by the ATT (2) or final (1) models. In comparison to conventional analysis of variance (ANOVA), the simulation approach provided similar point estimates and confidence intervals of exposure ratios. This PopPK based approach can be considered as a method of choice in situations where the simulation of HP data would not be an extrapolation, and when no other complicating factors are present.

Employing zero-inflated beta distribution in an exposure-response analysis of TYK2/JAK1 inhibitor brepocitinib in patients with plaque psoriasis.

Brepocitinib is an oral selective dual TYK2/JAK1 inhibitor and based on its cytokine inhibition profile is expected to provide therapeutic benefit in the treatment of plaque psoriasis. Efficacy data from a completed Phase 2a study in patients with moderate-to-severe plaque psoriasis were utilized to develop a population exposure-response model that can be employed to inform dose selection decisions for further clinical development. A modeling approach that employs the zero-inflated beta distribution was used to account for the bounded nature and distributional characteristics of the Psoriasis Area and Severity Index (PASI) score data. The developed exposure-response model provided an adequate description of the observed PASI scores across all the treatment arms tested and across both the induction and maintenance dosing periods of the study. In addition, the developed model exhibited a good predictive capacity with regard to the derived responder metrics (e.g., 75%/90%/100% improvement in PASI score [PASI75/90/100]). Clinical trial simulations indicated that the induction/maintenance dosing paradigm explored in this study does not offer any advantages from an efficacy perspective and that doses of 10, 30, and 60 mg once-daily may be suitable candidates for clinical evaluation in subsequent Phase 2b studies.

The ex vivo pharmacology of HIV-1 antiretrovirals differs between macaques and humans.

Non-human primates (NHP) are widely used for the pre-clinical assessment of antiretrovirals (ARVs) for HIV treatment and prevention. However, the utility of these models is questionable given the differences in ARV pharmacology between humans and macaques. Here, we report a model based on ex vivo ARV exposure and the challenge of mucosal tissue explants to define pharmacological differences between NHPs and humans. For colorectal and cervicovaginal explants in both species, high concentrations of tenofovir (TFV) and maraviroc were predictive of anti-viral efficacy. However, their combinations resulted in increased inhibitory potency in NHP when compared to human explants. In NHPs, higher TFV concentrations were measured in colorectal versus cervicovaginal explants (p = 0.042). In humans, this relationship was inverted with lower levels in colorectal tissue (p = 0.027). TFV-resistance caused greater loss of viral fitness for HIV-1 than SIV. This, tissue explants provide an important bridge to refine and appropriately interpret NHP studies.

Influence of excess ligand on Nephrogenic Systemic Fibrosis associated with nonionic, linear gadolinium-based contrast agents.

BACKGROUND: The molecular structure, charge, thermodynamic and kinetic stability are approximately the same for gadodiamide and gadoversetamide, the main substantive difference is that gadodiamide is manufactured with 5% free ligand to form Omniscan® and gadoversetamide with 10% free ligand to form OptiMARK®. PURPOSE: To determine the relative risk of Nephrogenic Systemic Fibrosis (NSF) between gadodiamide (Omniscan®) and gadoversetamide (OptiMARK®) and to explore the potential contribution of the amount of excess ligand added to their commercial formulations. MATERIALS AND METHODS: In this retrospective observational study, the number of doses and NSF cases associated with these agents were calculated based on two different approaches: the number of doses was determined based on pharmaceutical companies' information, and the number of unconfounded NSF cases was obtained from the previously published literature based on a legal database. A second analysis estimates the number of doses and NSF cases from the Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS). RESULTS: Approximately 87 million and 12 million doses of Omniscan® and OptiMARK®, respectively, have been administered worldwide since their original approval for use in the various countries throughout the world. A total of 197 and 8 unconfounded cases of NSF have been reported with Omniscan® and OptiMARK®, rendering an incidence of 2.3/million and 0.7/million for these agents, respectively. The FAERS analysis suggested reported incidences of 13.1/million and 5.0/million. CONCLUSION: There is an approximately 3-fold greater incidence of NSF from Omniscan® than OptiMARK®. The difference in incidence might reflect the lesser quantity of added free ligand to the formulation of Omniscan®.

Impact of chelation timing on gadolinium deposition in rats after contrast administration.

OBJECTIVE: To determine if gadolinium (Gd) can be rechelated once released from Gd-based contrast agents (GBCAs) and deposited in vivo. Despite extensive research comparing GBCAs and GBCA formulations as well as the ongoing debate about their risks of deposition and the role of Gd release, it remains unknown if retained Gd can be eliminated by administering chelating agents. MATERIALS AND METHODS: Rats were injected intravenously with 10 doses of 1 mmol/kg gadodiamide and treated with intravenous Zn-DTPA (30 µmol/kg) concomitantly or 1, 4 or 8 h after GBCA administration (N = 3 rats per group). After euthanization, tissues were harvested three days after the last dose of gadodiamide and tissue Gd concentrations were assessed by ICP-MS. Additionally, a simulation of a single 0.1 mmol/kg gadopentetate dose with 30 µmol/kg DTPA given either concomitantly or within the first 24 h after GBCA was run; simulated tissue Gd concentrations were compared with those observed in rats to determine if simulated trends were accurate. RESULTS: Concomitant DTPA did not produce a significant reduction in Gd concentration in any organ for rats. There was a time-dependent trend in liver Gd reduction. The 1 h timepoint was associated with a non-significant increase in kidney, brain and femur Gd relative to untreated controls. There were no significant deviations from the model-predicted Gd changes. DISCUSSION: Both the simulation and rat study did not identify major benefits for chelation at the doses given, despite the simulation assuming all Gd deposited in tissues is unchelated. The potential redistribution in the rat study provide a compelling result that may impact the clinical relevance of further work investigating rechelation of Gd. Future work should further describe the three-dimensional dose-time-response relationship for preventing Gd deposition, and how that relates to long-term Gd toxicities.

The Impact of Excess Ligand on the Retention of Nonionic, Linear Gadolinium-Based Contrast Agents in Patients With Various Levels of Renal Dysfunction: A Review and Simulation Analysis.

The role of gadolinium (Gd)-based contrast agents (GBCAs) in the pathophysiology of nephrogenic systemic fibrosis (NSF) is now uncontested. Although the definitive mechanism has not been established, the association with weaker GBCA ligands and with reduced renal clearance supports a hypothesis that Gd release from the GBCAs is a key process in precipitating the disease. Prevention strategies often include the use of more stable GBCA ligands in patients with reduced kidney function, but animal models and some clinical data suggest that better patient outcomes can be achieved when excess ligand is administered with weaker GBCAs; this is particularly significant for OptiMARK, which contains a nonionic, linear ligand similar to gadodiamide, the active ingredient in Omniscan, but contains twice the amount of excess ligand. Here we review evidence regarding the use of OptiMARK over Omniscan for prevention of NSF and perform a pharmacokinetic-based simulation to determine if the presented evidence is consistent with the established kinetics of GBCAs and Gd.

The stability of gadolinium-based contrast agents in human serum: A reanalysis of literature data and association with clinical outcomes.

PURPOSE: To reanalyze literature data of gadolinium (Gd)-based contrast agents (GBCAs) in plasma with a kinetic model of dissociation to provide a comprehensive assessment of equilibrium conditions for linear GBCAs. METHODS: Data for the release of Gd from GBCAs in human serum was extracted from a previous report in the literature and fit to a kinetic dissociation/association model. The conditional stabilities (logKcond) and percent intact over time were calculated using the model rate constants. The correlations between clinical outcomes and logKcond or other stability indices were determined. RESULTS: The release curves for Omniscan®, gadodiamide, OptiMARK®, gadoversetamide Magnevist® and Multihance® were extracted and all fit well to the kinetic model. The logKconds calculated from the rate constants were on the order of ~4-6, and were not significantly altered by excess ligand or phosphate. The stability constant based on the amount intact by the initial elimination half-life of GBCAs in plasma provided good correlation with outcomes observed in patients. CONCLUSIONS: Estimation of the kinetic constants for GBCA dissociation/association revealed that their stability in physiological fluid is much lower than previous approaches would suggest, which correlates well with deposition and pharmacokinetic observations of GBCAs in human patients.

A Strategy for Dosing Vancomycin to Therapeutic Targets Using Only Trough Concentrations.

Effective treatment of complicated methicillin-resistant Staphylococcus aureus (MRSA) infections with vancomycin requires a 24-h area under the concentration-time curve (AUC24) to minimum inhibitory concentration (MIC) ratio of at least 400. To ensure goal AUC24 has been reached requires either dosing to concentrations strongly associated with nephrotoxicity, measurement of patient-specific pharmacokinetics, or use of Bayesian statistics. In this study, we show a method of determining patient-specific pharmacokinetics and dosing to therapeutic AUC24 while minimizing potentially toxic concentrations, guided by only trough measurements. A Monte-Carlo simulation of 10,000 patients with complicated MRSA infections was prepared from two-compartment pharmacokinetic parameters using patient data extracted from the literature. The proposed method of determining patient-specific pharmacokinetics using consecutive trough concentrations was found to be more accurate than the conventional peak-trough method for peaks measured up to 4 h after infusion. Simulated human error in trough timing was found to reduce accuracy of the consecutive trough method, but an approach to resolve timing errors during a loading sequence or at steady-state using iteration is proposed. Both the simulated minimized concentration strategy and trough-based dosing to 15-20 mg/L had a high probability of achieving AUC24 at least 400 mg·h/L, but conventional trough-based dosing was associated with higher probability of potentially toxic 24-h doses and trough concentrations. The proposed strategy must be validated in real patients, with outcomes assessed before it is used in daily practice, but the theoretical benefits found in the simulation suggest this simple strategy should be considered with other AUC24-based approaches.

Cross-Reactivity of Select PEG-Binding Antibodies to Other Polymers Containing a C-C-O Backbone.

Polyethylene glycol (PEG), a flexible and relatively hydrophilic polymer, is widely used not only in medicine but also in numerous everyday hygiene, food, and skincare products. Recent animal and human studies have shown that antibodies (Abs) that bind PEG can be induced, leading to markedly reduced therapeutic efficacy of PEGylated therapeutics as well as possibly resulting in acute anaphylaxis and hypersensitivity reactions. Because humans are exposed to numerous other synthetic polymers, we sought to investigate whether such "anti-PEG" antibodies may also bind other synthetic polymers, particularly those with structural similarities to PEG. In a screen of six commercially available and two recombinantly produced anti-PEG IgG and IgM antibodies, we found five antibodies (3 IgG and 2 IgM) that readily bind polypropylene glycol (PPG), polytetramethylene ether glycol (PTMEG), and poly-1,4-butylene adipate (PBA). In contrast, none of the eight antibodies bound dextran (DEX) or polyepoxysuccinic acid (PES), and only two exhibited detectable affinity to polyethylenimine (PEI), suggesting that these PEG-binding antibodies likely possibly recognizable accessible C-C-O groups in the polymer backbone. We also observed similar cross-reactivity in plasma of human subjects with high titers of PEG-binding IgG and IgM. These results directly demonstrate potential cross-reactivity of select PEG-binding antibodies, which represents a new category of antidrug antibodies whereby an adverse immune response can be elicited as a result of prior exposures to PEG or other synthetic PEG-like polymers.

Gadolinium deposition in the brain: Lessons learned from other metals known to cross the blood-brain barrier.

The recent discovery of gadolinium (Gd) deposition in the brains of patients receiving Gd-based contrast agents (GBCAs) raises several important questions including by what mechanism Gd or GBCAs pass through the blood-brain barrier. Decades of research focused on the safety and stability of GBCAs have not identified any mechanism of uptake. Here we review findings of Gd deposition from human and animal data, and how distribution mechanisms elucidated for endogenous and toxic metals may explain entrance of Gd into the central nervous system. Three general uptake mechanisms are considered along with examples of metals known to enter the central nervous system by these routes: (1) carrier-mediated, (2) transporter-mediated and (3) passive. The potential for chelation therapy to reduce deposition is also discussed. The work reported for other metals provides guidance for how the mechanism of Gd deposition in the brain can be determined which is essential information for rational prevention or treatment.

Can gadolinium be re-chelated in vivo? Considerations from decorporation therapy.

Gadolinium (Gd) and Gd-based contrast agents (GBCAs) have been observed to deposit in tissues of patients following contrast enhanced MR imaging procedures. A conservative approach for chelation therapy of this toxic metal dictates the assumption that minimal intact GBCAs are present. Currently the extent to which these deposits are primarily de-chelated Gd remains uncertain, prevailing knowledge suggests that for linear agents much of the Gd is de-chelated, while for the macrocyclic agents, the Gd may be still largely chelated. To extract Gd from tissues and facilitate its release, chelation therapy should be both safe and effective. Here we discuss chelation therapy as it relates to Gd deposition. The principles of chelation are reviewed, initially with reference to ligand stability in complex biological fluids. A model of decorporation and how it relates to elimination of Gd deposits is also reviewed. When more is learned about Gd deposition, optimal removal strategies must be developed using basic thermodynamic and kinetic principles.

Vancomycin Trough Concentration as a Predictor of Clinical Outcomes in Patients with Staphylococcus aureus Bacteremia: A Meta-analysis of Observational Studies.

STUDY OBJECTIVE: To determine the strength of evidence for better clinical outcomes in patients with Staphylococcus aureus bacteremia who had vancomycin trough levels of 15-20 mg/L. DESIGN: Meta-analysis of 14 observational cohort studies. PATIENTS: A total of 1677 patients, representing geriatric and unspecified inpatients, who received standard dosing of vancomycin for the treatment of S. aureus bacteremia and who had trough level goals of 15-20 mg/L. MEASUREMENTS AND MAIN RESULTS: The treatment variables examined in the analysis were vancomycin trough concentrations and 24-hour area under the concentration-time curve to minimum inhibitory concentration ratio (AUC:MIC) values. The outcomes of interest were mortality, persistent bacteremia, and treatment failure. Mortality was defined as 30-day mortality, in-hospital mortality, or a comparable measure; persistent bacteremia was defined as bacteremia lasting at least 7 days after the initiation of vancomycin; treatment failure was defined as a composite end point that included at least persistent bacteremia and mortality, as previously defined. Higher vancomycin trough levels (15 mg/L or greater or based on MIC) were not associated with significantly reduced treatment failure, persistent bacteremia, or mortality. Higher AUC:MIC values were associated with significantly reduced treatment failure (odds ratio [OR] 0.41, 95% confidence interval [CI] 0.31-0.53), persistent bacteremia (OR 0.53, 95% CI 0.33-0.86), and mortality (OR 0.47, 95% CI 0.33-0.65). The weighted mean ± SD AUC:MIC threshold defined by regression analyses in the included studies was 418 ± 88 hours, which supports the current goal of 400 hours or more. CONCLUSION: Vancomycin trough concentrations do not have sufficient evidence to support their use as the primary guide in vancomycin dosing. Dosing should instead focus on AUC:MIC values, which have strong evidence of benefit.

Flux through silicone and human skin fitted to a series/parallel model.

BACKGROUND: Recent reports of the good correlation between maximum flux through human skin in vitro from water, JMHAQ, and maximum flux through silicone from water, JMPAQ, demand that the mechanism of maximum flux across these two apparently quite different membranes be compared to understand the bases of the correlation. RESULTS/DISCUSSION: A n = 70 log JMPAQ database and a matched n = 55 log JMHAQ database of molecules were found to fit well to a series/parallel model where three parallel solubility dependent pathways existed: a lipid pathway, an aqueous pathway, and a series pathway of alternating lipid and aqueous phases. CONCLUSION: The results of this analysis surprisingly suggest that the architecture of the two membranes present similar solubility based pathways through which drugs diffuse.

The Flux of Phenolic Compounds through Silicone Membranes.

Phenols as a class of molecules have been reported to exhibit higher log maximum fluxes through human stratum corneum, SC, from water, log JMHAQ, than other classes of molecules. This suggests that their corresponding log maximum fluxes through silicone from water, log JMPAQ, may be useful to extend the existing n = 63 log JMPAQ database to include more log JMPAQ values greater than 0.0. The log JMPAQ values for n = 7 phenols predicted to give log JMPAQ values greater than 0.0 based on their log JMHAQ values have been experimentally determined. These n = 7 new log JMPAQ values have been added to the existing n = 63 log JMPAQ database to give a new n = 70 database and the n = 7 literature log JMHAQ values have been added to the existing n = 48 log JMHAQ database (matched to the n = 63 log JMPAQ database) to give a new n = 55 database. The addition of the n = 7 phenols improved the correlations of these flux databases when fitted to the Roberts-Sloan equation, RS, as well as the correlation between the matched experimental (Exp.) log JMPAQ with the Exp. log JMHAQ.