Liver–Heart Co-culture Links Terfenadine Clearance with Contractile Recovery
A coupled human liver–heart system followed terfenadine, its metabolite and cardiac motion over time. Contractile recovery occurred alongside hepatic metabolism, while material adsorption and exposure level shaped the experiment.
Measuring Metabolism and Cardiac Function Together
Terfenadine provides a useful case for studying how hepatic metabolism changes a drug's effect on the heart. Researchers at Japan's National Institute of Health Sciences and collaborating universities combined primary human hepatocytes with engineered heart tissues to follow that relationship over time. The study measured drug and metabolite concentrations together with cardiac contraction in a shared circulation system.
The heart tissues were generated from human iPSC-derived cardiomyocytes with supporting stromal cells. They were cultured in a plate with an integrated stirring pump and a bottom oxygen-permeable membrane, alongside a separate hepatocyte compartment. The comparison included heart tissue alone, liver cells alone and the coupled liver–heart configuration. This arrangement allowed metabolism-associated changes to be compared with the cardiac response under matched nominal exposure conditions.
A Dose- and Time-Dependent Functional Response
Terfenadine was introduced at nominal concentrations of 1 or 3 μM. Cardiac motion and culture-medium samples were collected at 4, 24 and 48 hours. Video analysis quantified contraction and relaxation, while LC–MS/MS measured terfenadine and fexofenadine.
At 1 μM, the co-cultured heart tissues initially showed marked functional suppression, with contraction at approximately 6% of the pre-exposure baseline at four hours. By 24 hours, contraction had recovered to approximately 70% in co-culture, compared with approximately 8% in heart-tissue monoculture. At 48 hours, the coupled tissues had returned to their pre-exposure contraction level, while the monocultures reached approximately 30%.
The higher concentration produced a different course. At 3 μM, heart-tissue monocultures remained in a fibrillation-like state through 48 hours. Co-cultures showed partial recovery to approximately 30% of baseline at that time, with irregular beating periods and amplitudes. Similar time- and exposure-dependent patterns were observed in contraction and relaxation velocities.
Following the Parent Drug and Metabolite
Chemical analysis showed loss of terfenadine and formation of fexofenadine in the hepatocyte-containing systems. Metabolite production was nearly undetectable in heart-tissue monoculture. These measurements link the functional comparison with the metabolic capability supplied by the liver compartment.
The timing also contained information beyond the final concentration. At the lower exposure, terfenadine fell below the analytical detection limit by 24 hours in co-culture, yet cardiac function had not fully normalized. The delay between extracellular drug loss and complete contractile recovery illustrates why time-resolved measurements can be useful in a coupled organ model.
The authors discuss reversible ion-channel interactions as a possible explanation for the recovery pattern. That explanation remains a mechanistic interpretation of the observed concentration and motion data. The experiment reports a temporal pharmacokinetic–pharmacodynamic relationship rather than a fully fitted quantitative PK–PD model.
Accounting for the Experimental Exposure
A cell-free comparison showed that the PDMS supports used for the heart tissues could remove terfenadine from the medium by adsorption. Consequently, the initial nominal concentration and the measured concentration experienced by the tissues differed. Measuring exposure in each configuration is important when interpreting the apparent pace of clearance and the resulting functional response.
The researchers deliberately used concentrations that produced a clear pharmacodynamic signal for this proof-of-concept experiment. They also used a single hepatocyte batch and a small number of independent tissue constructs. These conditions support analysis of this particular metabolism-associated response, while broader drug and donor comparisons would be needed to establish how the platform performs across mechanisms.
The practical result is an experimental framework that retains chemistry and tissue function on the same time axis. Its value lies in resolving the sequence of functional suppression, parent-drug depletion, metabolite formation and recovery, including the imperfect correspondence between those processes.