Copanlisib population pharmacokinetics from phase I-III studies and exposure-response relationships in combination with rituximab

Copanlisib dose selection was established under the maximum tolerated dose paradigm, and no dedicated dose-finding studies have investigated copanlisib dose selection when used in combination with rituximab. In CHRONOS-3, copanlisib plus rituximab demonstrated significantly improved progression-free survival versus placebo plus rituximab in patients with relapsed indolent non-Hodgkin lymphoma (iNHL). We conducted a comprehensive investigation of copanlisib population pharmacokinetics (PopPK) from a pooled analysis of 712 patients across nine copanlisib phase I-III studies and exposure-response (ER) relationships for efficacy and safety from the 1-year follow-up of CHRONOS-3. PopPK analyses examined the impact of demographic, laboratory, and comedication covariates on copanlisib between-patient PK variability. Individual static and time-varying exposure estimates were derived to investigate exposure-efficacy and exposure-safety relationships. Multivariate Cox proportional hazards and logistic regression analyses examined ER relationships with consideration of predefined potentially prognostic demographic-, laboratory-, and/or disease-related baseline covariates. Copanlisib PK were best described by a three-compartment model with first-order elimination. Individual identified covariates had modest effects on copanlisib PK and were generally in line with known copanlisib disposition properties. In CHRONOS-3, ER analyses showed a significant relationship between time-varying exposure estimates and progression-free survival, and no significant exposure-safety relationships. Thus, lower copanlisib doses may result in reduced efficacy but not necessarily improved safety or tolerability. These outcomes substantiate the current intermittent dosing regimen of copanlisib 60 mg on days 1, 8, and 15 of a 28-day cycle and support the observed clinical results of copanlisib in combination with rituximab in the iNHL population.

Abstract A25: Copanlisib population pharmacokinetics (popPK) and exposure-response (ER) in relapsed indolent non-Hodgkin lymphoma (iNHL): Analysis from the 1-year follow-up of the phase III, randomized CHRONOS-3 trial

Introduction: CHRONOS-3 randomized patients (pts) with relapsed iNHL to the PI3K inhibitor copanlisib 60 mg administered on Day 1, 8, 15 of 28-day cycle in combination with standard rituximab (C+R) or placebo plus rituximab (P+R) and demonstrated superior progression-free survival (PFS) of C+R vs. P+R (hazard ratio[HR]: 0.52; Matasar et al. Lancet Oncol 2021). In an updated 1-year follow-up analysis, the benefit of C+R was confirmed in the overall relapsed iNHL population (Zinzani et al., EHA 2022) and in the subset of pts with FL (see Capra et al. AACR Lymphoma 2022 abstract). We investigated popPK and ER relationships for copanlisib efficacy and safety from a pooled analysis, with a particular focus on the 1-year follow-up of CHRONOS-3 to confirm copanlisib dose selection. Methods: A comprehensive popPK model for copanlisib was developed using plasma concentrations from 712 pts across 9 copanlisib clinical Phase 1-3 studies. Demographic, laboratory, and comedication covariates were investigated for their influence on copanlisib between-patient PK variability. Individual static and time-varying exposure estimates for pts enrolled in CHRONOS-3 were derived to investigate relationships to efficacy (PFS, overall response rate [ORR], complete response [CR]) and key treatment-emergent safety events using multivariate cox proportional hazards (CPH) and logistic regression analyses, after accounting for predefined potentially prognostic demographic, laboratory, and/or disease related baseline covariates. Results: Copanlisib PK was best described by a 3-compartment PK model with first-order elimination from the central compartment following intravenous infusion. Copanlisib PK was dose proportional and time independent with no accumulation. Covariates statistically selected as influencing copanlisib PK included rifampin (strong CYP3A inducer) and itraconazole (strong CYP3A inhibitor) co-administration, female sex, Japanese pts, mild or worse hepatic impairment, and a CHRONOS-3 study effect, however all identified covariates, aside from rifampin/itraconazole, had no clinically relevant effect (≤ 35%) on copanlisib AUC. In CHRONOS-3, CPH analyses demonstrated a statistically significant positive ER relationship for PFS where greater copanlisib exposure was associated with prolonged PFS. Relative to P+R, the HR for C+R at the 5th and 95th percentile of copanlisib time-varying exposure was 0.875 and 0.217, respectively. Thus, lower copanlisib doses may result in reduced efficacy in relapsed iNHL. CPH and/or logistic regression analyses demonstrated no significant ER relationship for ORR and investigated adverse events. Based on the modelling, lower starting doses of copanlisib are not anticipated to improve safety and tolerability. The ER analyses support the overall positive benefit/risk assessment of copanlisib reported from the clinical efficacy and results of CHRONOS-3 in iNHL. Conclusions: PopPK and ER analyses support copanlisib 60 mg administered on Day 1, 8, 15 of 28-day cycle dose selection in the relapsed iNHL population.

Citation Format: Jonathan Moss, Dirk Garmann, Rupert Austin, Florian Hiemeyer, Patricia DeLora, Matthew J Matasar, Pier Luigi Zinzani, Vita Beckert, Lidia Mongay Soler, Barrett H Childs, Peter N Morcos. Copanlisib population pharmacokinetics (popPK) and exposure-response (ER) in relapsed indolent non-Hodgkin lymphoma (iNHL): Analysis from the 1-year follow-up of the phase III, randomized CHRONOS-3 trial [abstract]. In: Proceedings of the Third AACR International Meeting: Advances in Malignant Lymphoma: Maximizing the Basic-Translational Interface for Clinical Application; 2022 Jun 23-26; Boston, MA. Philadelphia (PA): AACR; Blood Cancer Discov 2022;3(5_Suppl):Abstract nr A25.

Predicting menstrual bleeding patterns with levonorgestrel-releasing intrauterine systems

PURPOSE

To develop a bleeding-pattern prediction model to inform counselling on amount and regularity of bleeding after levonorgestrel-releasing intrauterine system (LNG-IUS) placement.

MATERIALS AND METHODS

Fixed-cluster and regression-tree models were developed using bleeding data pooled from two clinical trials of LNG-IUSs. Models were trained and cross-validated on LNG-IUS 12 data, then applied to LNG-IUS 20 and LNG-IUS 8 data. Three clusters were generated for the fixed-cluster model: predominantly amenorrhoea; predominantly spotting; and predominantly bleeding. A random-forest model predicted the future-bleeding cluster, then the probability of cycle regularity was calculated. In the regression-tree model, women were assigned by the model to less- or more-bleeding groups.

RESULTS

With LNG-IUS 12 (n = 1351) in the fixed-cluster model, 70.4% of women were correctly classified. The correct classification rates for LNG-IUS 20 (n = 216) and LNG-IUS 8 (n = 1300) were 72.2% and 69.0%. The probability distribution for cycle regularity showed regular and irregular bleeding were best separated with LNG-IUS 12 data, and less well with LNG-IUS 20 and LNG-IUS 8 data. In the regression-tree model there was high variability in the more- and less-bleeding group distributions with LNG-IUS 12 data.

CONCLUSIONS

A fixed-cluster model predicted bleeding patterns better than a regression-tree model in women using LNG-IUS, yielding understandable, informative output.

Insertion experience of women and health care professionals in the Kyleena® Satisfaction Study

Purpose: The Kyleena^® Satisfaction Study (KYSS) is a prospective, observational study conducted to assess satisfaction with LNG-IUS 12 (Kyleena^®) in clinical practice and aims to provide adequate information for counselling women on what to expect regarding insertion and satisfaction.Materials and methods: Women deciding to use LNG-IUS 12 during routine counselling were informed of the study and provided informed consent. A baseline analysis was conducted to evaluate demographics, ease of insertion assessed by investigators, pain at insertion rated by women, additional interventions for insertion, and adverse events.Results: 1,110 women (536 parous, 574 nulliparous) had an insertion attempt and were included. Insertion was rated as easy in 494 (92.2%) parous and 516 (89.9%) nulliparous women. Pain was assessed as none or mild by 475 (88.6%) parous and 387 (67.4%) nulliparous women. Additional interventions were not required for most insertions (705; 63.6%). Overall 111 (10.0%) women reported adverse events at the time of baseline analysis.Conclusions: This analysis demonstrates that LNG-IUS 12 insertion is easy and associated with no or mild pain in most women. Additional interventions for insertion are not required in most cases. After 3 months, the number of adverse events is low.Implications: The present baseline analysis of the Kyleena^® Satisfaction Study (KYSS) demonstrates that most women rate insertion pain of LNG-IUS 12 as none or mild and clinicians consider insertion easy in the majority of cases.

Bleeding patterns with the 19.5 mg LNG-IUS, with special focus on the first year of use: implications for counselling

Objective: The aim of the study was to provide an additional, detailed description of early bleeding patterns with the 19.5 mg levonorgestrel-releasing intrauterine system (LNG-IUS). Methods: We conducted a pooled analysis of the bleeding diaries of participants in a previously reported phase II randomised controlled study (n = 741) and a phase III study (n = 2904), with 2-year extension phase (n = 707), of the 19.5 mg LNG-IUS. Main outcome measures were the median number of bleeding and/or spotting days per 30-day reference period for 12 months and the influence of the previous contraceptive method and levonorgestrel dose on bleeding patterns. Results: The pooled analysis comprised 1697 women. There was a progressive decline in the number of bleeding and/or spotting days from month 1: the proportion of women with ≤4 bleeding and/or spotting days per month increased from 6.2% in month 1 to 15.8% in month 2, 26.0% in month 3, 39.3% in month 6 and 54.1% in month 12. The median number of bleeding and/or spotting days in month 1 was lowest in women who had previously been using an LNG-IUS. Conclusion: Analysis of bleeding diaries using 30-day reference periods provides detailed insight into bleeding changes in the first months following placement of the 19.5 mg LNG-IUS. This insight may prove useful when counselling women about contraceptive choice and method continuation.

Impact of the presequence of a mitochondrium-targeted precursor, preadrenodoxin, on folding, catalytic activity, and stability of the protein in vitro

Bovine preadrenodoxin, an adrenocortical precursor protein destined for mitochondrial import, was expressed in Escherichia coli as an [2Fe-2S] cluster-containing protein. It was found in inclusion bodies, purified from there, and finally reconstituted to obtain soluble holo-protein. The impact of the presequence on folding of the protein using biochemical and biophysical approaches has been investigated. Upon unfolding the preprotein reveals a decrease in the denaturational enthalpy and heat capacity compared with mature adrenodoxin, indicating an incomplete unfolding of the preprotein with remaining residual structure. Moreover, the data obtained show that the presequence is solvent exposed in aqueous solution with no preference for secondary structure elements and that it does not disturb the accurate folding of the mature part of the protein. The latter conclusion is also based on the finding that the precursor in vitro exhibits electron transfer function comparable to the mature protein, adrenodoxin. While the reduction of cytochrome c, reflecting the interaction between adrenodoxin and its reductase, and the interaction with CYP11B1 have not been significantly affected by the presence of the presequence, the binding affinity of preadrenodoxin to CYP11A1 is 5.5-fold lower than that of the mature form.

Cluster-iron substitution is related to structural and functional features of adrenodoxin mutants and to their redox states

Site-directed mutants of adrenodoxin were studied for their ability to undergo cluster-iron substitution when reacted with zinc or cadmium salts under non-denaturing conditions in the presence or absence of reductants. Equilibrium and kinetic data for metal substitution were correlated with data on the stability to thermal unfolding and with the redox potential of the protein. Similarly to the wild-type protein, all mutants were able to stabilize a substituted form of the protein containing two metal (Zn or Cd) atoms and two sulfide ions/mol protein and a substituted form of the protein containing two sulfide ions and five Cd atoms/mol protein. However, the distribution of these two metal-substituted forms was different among the investigated proteins. [Ser95]Adrenodoxin stabilized either metal-substituted forms, confirming that Cys95 is not involved in metal coordination, even when five Cd atoms are bound to the protein. Removal of the extremely conserved hydroxy function at position 54 resulted in complete apoprotein formation upon reaction with Cd (75 % with Zn) under reducing conditions, indicating a cluster-harboring role for this function, which is conserved in all known 2Fe-2S proteins. Mutants at His56, which represents a residue unique to most vertebrate-type ferredoxins, were much more reactive than the wild-type protein with either metal, indicating that His56 plays a prominent role in the stabilization of the protein structure in the immediate vicinity of the cluster in this class of proteins. The nature of the metal-substitution products was dependent on cluster accessibility. For the reduced proteins, apoprotein formation depended on protein stability, while the velocity of metal substitution depended on the ease of cluster reduction.

Specific aspects of electron transfer from adrenodoxin to cytochromes p450scc and p45011beta

An analysis of the electron transfer kinetics from the reduced [2Fe-2S] center of bovine adrenodoxin and its mutants to the natural electron acceptors, cytochromes P450scc and P45011beta, is the primary focus of this paper. A series of mutant proteins with distinctive structural parameters such as redox potential, microenvironment of the iron-sulfur cluster, electrostatic properties, and conformational stability was used to provide more detailed insight into the contribution of the electronic and conformational states of adrenodoxin to the driving forces of the complex formation of reduced adrenodoxin with cytochromes P450scc and P45011beta and electron transfer. The apparent rate constants of P450scc reduction were generally proportional to the adrenodoxin redox potential under conditions in which the protein-protein interactions were not affected. However, the effect of redox potential differences was shown to be masked by structural and electrostatic effects. In contrast, no correlation of the reduction rates of P45011beta with the redox potential of adrenodoxin mutants was found. Compared with the interaction with P450scc, however, the hydrophobic protein region between the iron-sulfur cluster and the acidic site on the surface of adrenodoxin seems to play an important role for precise complementarity in the tightly associated complex with P45011beta.

Conformational stability of adrenodoxin mutant proteins

Adrenodoxin and the mutants at the positions T54, H56, D76, Y82, and C95, as well as the deletion mutants 4-114 and 4-108, were studied by high-sensitivity scanning microcalorimetry, limited proteolysis, and absorption spectroscopy. The mutants show thermal transition temperatures ranging from 46 to 56 degrees C, enthalpy changes from 250 to 370 kJ/mol, and heat capacity change delta Cp = 7.28 +/- 0.67 kJ/mol/K, except H56R. The amino acid replacement H56R produces substantial local changes in the region around positions 56 and Y82, as indicated by reduced heat capacity change (delta Cp = 4.29 +/- 0.37 kJ/mol/K) and enhanced fluorescence. Deletion mutant 4-108 is apparently more stable than the wild type, as judged by higher specific denaturation enthalpy and resistance toward proteolytic degradation. No simple correlation between conformational stability and functional properties could be found.

Mutational effects on the spectroscopic properties and biological activities of oxidized bovine adrenodoxin, and their structural implications

Of the aromatic 1H-NMR signals of oxidized bovine adrenodoxin only those of His56 showed intrinsic chemical shift changes upon replacement of Tyr82 by Ser or Leu, that must arise from a loss of a through-space ring-current effect of the tyrosine ring in these mutants. Thus, of the three His residues contained in adrenodoxin, His56 is closest to Tyr82, and hence to the highly acidic determinant region of adrenodoxin that is the interaction site for adrenodoxin reductase and P-450. The strong dependence of the fluorescence intensity of Tyr82 on the residue in position 56 supported this observation. As a consequence of this, the effects of replacement of His56 by Gln or Thr on cytochrome c reduction and cytochromes P-450(11 beta) (CYP11B1)-dependent and P-450scc (CYP11A1)-dependent substrate conversions were studied. No influence on Vmax values was observed for all reactions mediated by the mutants, implying His56 does not play a decisive role in the intramolecular or intermolecular electron transfer. In contrast, the Km values were increased, as was the Ks value for binding of CYP11A1 to the [H56T]adrenodoxin. The secondary structure deduced from further NMR data of adrenodoxin was compared with that of other ferredoxins. Tyr82 is in a region of the molecule containing no secondary-structure elements. The data for Tyr82 are in keeping with the biological activities and suggests it is in a flexible, solvent-exposed region of the molecule.

Immunocytochemical localization of heterologously expressed adrenodoxin and its electron acceptor cytochrome P45011B1 in Escherichia coli

Adrenal steroid hydroxylase P45011B1 and its electron donor adrenodoxin were localized in the cortex of bovine adrenals by immunogold-silver staining. In order to test recently developed heterologous expression systems for both enzymes to enable structure-function studies, immunocytochemical marker methods were applied. Adrenodoxin, the ferredoxin of the adrenal gland, was successfully expressed and for the first time localized in Escherichia coli. By use of ultrathin cryosections and the protein A-gold technique, adrenodoxin was detectable in large amounts in the cytoplasm of the bacterial cells, and, following the insertion of the outer membrane protein A leader sequence of E. coli, also in the periplasmic space. A fusion protein between mature adrenodoxin and human P45011B1 was constructed and clearly localized in E. coli by antibodies against both proteins.

Mutations of tyrosine 82 in bovine adrenodoxin that affect binding to cytochromes P45011A1 and P45011B1 but not electron transfer

To understand the function of the unique tyrosine in position 82 of bovine adrenodoxin (Adx), which had been proposed to be involved in electron transfer from NADPH-dependent adrenodoxin reductase (AdR) to cytochrome P-450 enzymes and/or AdR binding by chemical modification studies (Taniguchi, T., and Kimura, T. (1975) Biochemistry 14, 5573-5578), the residue was replaced by phenylalanine, leucine, or serine. Unchanged absorption, CD, and electron spin resonance spectra as well as redox potentials indicate that the environment of the [2Fe-2S] cluster was not affected by the mutations. The Vmax values in cytochrome c reduction, P45011A1- and P45011B1-dependent activities were also not changed when using Y82F, Y82S, and Y82L Adx mutants as electron donor, demonstrating that tyrosine 82 is not involved in the intra- or intermolecular electron transfer. Replacement of tyrosine 82 did not affect AdR binding as shown by unchanged cytochrome c activity. There are, however, changes in Km values up to 4-fold when measuring the enzymatic activities of mutant Adx with P45011A1 and P45011B1. These changes differ in dependence on the P-450 (P45011A1 or P45011B1) used. The results suggest that mutation of tyrosine 82 either directly or indirectly (by inducing small conformational changes of the binding domain) affects the binding of cytochromes P-450.

Expression of bovine adrenodoxin in E. coli and site-directed mutagenesis of /2 Fe-2S/ cluster ligands

Expression systems for adrenodoxin into the periplasm and the cytoplasm of E. coli have been developed as a prerequisite for site-directed mutagenesis studies. In both systems the /2Fe-2S/ cluster of the protein was correctly assembled, the cytoplasmic one gives, however, a tenfold higher expression level. To determine which of the five cysteines at positions 46, 52, 55, 92, and 95 coordinate the /2Fe-2S/ center, they have been individually mutated into serines. From these mutants, only C95S forms a functionally active holoprotein. Thus, residues 46, 52, 55, and 92 are the cysteines that coordinate the /2Fe-2S/ cluster in adrenodoxin.