Source and numerical evidence
1. Astemizole — Effect Predicted
The audited model output captures the reported ap prolongation; ead-like events.
Higher experimental doses have unavailable durations because of events; agreement uses eligible observations. EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
2. Azimilide — Effect Predicted
The audited model output captures the reported shift towards ap prolongation.
Low-dose response differs between cell lines; higher optical durations are frequently unavailable.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Shift towards AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
3. Bepridil — Effect Predicted
The audited model output captures the reported ap prolongation; heterogeneous response.
Dose response is heterogeneous; source high-dose quiescence is not a complete matched phenotype.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation; heterogeneous response
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
4. Chlorpromazine — Effect Predicted
The audited model output captures the reported ap prolongation in some assay series.
Prolongation was significant in only some laboratory datasets; the result does not describe every site.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation in some assay series
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
5. Cisapride — Effect Predicted
The audited model output captures the reported ap prolongation.
High-dose events limit experimental duration measurements.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
6. Clarithromycin — Relevant Effect Captured
Later prolongation captured; initial shortening missed.
Initial shortening was not reproduced, although the high-dose prolonging effect agrees.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Initial shortening, then prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
7. Clozapine — Effect Predicted
The audited model output captures the reported shift towards ap prolongation; variable response.
Experimental responses vary between sites and cell lines; the result concerns the main prolonging trend.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Shift towards AP prolongation; variable response
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
8. Diltiazem — Relevant Effect Captured
Shortening through 1 µM captured; direction reverses incorrectly at 10 µM.
Shortening agrees through 1 µM; at 10 µM the model changes towards prolongation. Direction agrees at both source-matched nominal doses; magnitude differs. Honda used immature hiPS cells at room temperature and 0.1 Hz, versus adult ORd at 1 Hz and ten fast cycles. Potency is independently sourced from Li 2019. Peak voltage is not action potential amplitude; amplitude equals peak minus resting potential.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP shortening
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Honda 2011
Observed APD90 change (%) · 201B7 hiPS-CM ventricular-like single cells
APD90 shortening
applied bath concentration
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
9. Disopyramide — Relevant Effect Captured
Lower-dose prolongation captured; high-dose quiescence and mixed response are not fully reproduced.
Lower-dose prolongation agrees; quiescence and mixed surviving-well responses limit the high-dose comparison.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Prolongation; mixed high-dose response
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
10. Dofetilide — Effect Predicted
The audited model output captures the reported ap prolongation; ead-like events.
Only eligible source durations are displayed; event frequency has not been quantitatively validated. EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles. Table 1 contains measured adult-human ex vivo APD90, not published simulations. Pharm and CiPA vary channel inputs against the same experiment; they are not independent validations. Ten cycles differ from the published model’s 1500 prepaces; extracellular ions also differ. Highest-dose APD90 median includes 439/500 valid cells and excludes unavailable crossings. Source APD analysis excluded one trabecula with EAD at 0.1 uM. Numerical agreement and clinical safety are not asserted. Table 1 contains measured adult-human ex vivo APD90, not published simulations. Pharm and CiPA vary channel inputs against the same experiment; they are not independent validations. Ten cycles differ from the published model’s 1500 prepaces; extracellular ions also differ. Highest-dose APD90 median includes 438/500 valid cells and excludes unavailable crossings. Source APD analysis excluded one trabecula with EAD at 0.1 uM. Numerical agreement and clinical safety are not asserted.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
Barral 2025
Observed APD90 change (ms) · adult human ventricular trabeculae
APD90 prolongation
nominal; Table 2 says concentration was not bioanalytically measured
Source-derived observation records and per-study source notes in the audited original comparison
Barral 2025
Observed APD90 change (ms) · adult human ventricular trabeculae
APD90 prolongation
nominal; Table 2 says concentration was not bioanalytically measured
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
11. Domperidone — Effect Predicted
The audited model output captures the reported dose-related ap prolongation.
Small low-dose changes differ, while the main shift towards prolongation agrees.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Dose-related AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
12. Droperidol — Effect Predicted
The audited model output captures the reported ap prolongation.
High-dose experimental durations are often unavailable.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
13. Ibutilide — Effect Predicted
The audited model output captures the reported ap prolongation.
Experimental events limit higher-dose duration measurements.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
14. Loratadine — Relevant Effect Captured
Shortening relative to control and absence of prolongation are captured. All eight published site-2 cell-line means are negative, but the observed profile is non-monotonic and does not follow the model's progressively greater shortening. The earlier inconclusive label described the dose gradient, not missing simulation data.
Only the relative-to-control shortening / lack of prolongation is aligned. A monotonic experimental concentration response is not established. Clinical QTc provides separate context and is not substituted for cellular APD90c. No oral-dose to bath-concentration conversion is made.
Blinova et al. 2018 · International Multisite Study of Human-Induced Pluripotent Stem Cell-Derived Cardiomyocytes for Drug Proarrhythmic Potential Assessment
Baseline- and vehicle-controlled, rate-corrected APD90 (ddAPD90c) · VSO site 2; Cor.4U and iCell2 hiPSC cardiomyocytes
All eight selected cell-line/dose means show shortening relative to baseline and vehicle, with a non-monotonic concentration profile and source-supplied SEM/n.
0.00095, 0.003, 0.00949 and 0.03 µM
Raw multisite supplemental workbook; Drug_Name=Loratadine, site=2, Platform=CLY (optical VSO); Supplemental Table 2 nominal doses. Existing 20261010 numeric QA verified all selected rows.
Kosoglou et al. 2000 · Pharmacokinetics and electrocardiographic pharmacodynamics of loratadine with ketoconazole or cimetidine
Clinical QTc · Healthy volunteers; randomized crossover
No significant QTc prolongation despite increased plasma exposure during coadministration.
10 mg loratadine daily for 10 days, alone or with metabolic inhibitors
Abstract Results and Conclusions; electrocardiographic evaluation
Observed values by source and concentration
Model outputs by study and concentration
15. Metoprolol — Effect Predicted
The audited model output captures the reported ap prolongation at high concentrations.
This is a high-dose cellular effect; clinical beta-blockade is not modelled.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation at high concentrations
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
16. Mexiletine — Relevant Effect Captured
Initial shortening captured; later response and quiescence are not fully reproduced.
Initial shortening agrees; the full high-dose profile and source quiescence are not reproduced.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP shortening; high-dose quiescence
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
17. Nifedipine — Effect Predicted
The audited model output captures the reported ap shortening.
Direction agrees; magnitudes and high-dose quiescence differ.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP shortening
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
18. Nitrendipine — Effect Predicted
The audited model output captures the reported ap shortening.
Shortening agrees; one high-dose optical series has only one eligible observation.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP shortening
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
19. Ondansetron — Effect Predicted
The audited model output captures the reported ap prolongation.
Higher optical durations are unavailable; agreement uses eligible lower-dose observations.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
20. Pimozide — Relevant Effect Captured
Prolongation captured; EADs were absent at all modeled active doses.
One cell line has an initial negative change; the later shift towards prolongation agrees. EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Shift towards AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
21. Quinidine — Effect Predicted
The audited model output captures the reported ap prolongation and arrhythmic events; ead-like events.
Preferred optical APD has one eligible dose at 0.95 µM; separate MEA evidence supports the general prolonging effect. EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles. Table 1 contains measured adult-human ex vivo APD90, not published simulations. Pharm and CiPA vary channel inputs against the same experiment; they are not independent validations. Ten cycles differ from the published model’s 1500 prepaces; extracellular ions also differ. Highest-dose APD90 median includes 487/500 valid cells and excludes unavailable crossings. Positive dose trend agrees, but magnitude is strongly overpredicted (338.7 ms versus 37 ± 7 ms). This does not validate quantitative APD, combined-block mitigation or clinical safety. Only IKr and ICaL potency are represented; neutral effects on other channels are assumptions. Table 1 contains measured adult-human ex vivo APD90, not published simulations. Pharm and CiPA vary channel inputs against the same experiment; they are not independent validations. Ten cycles differ from the published model’s 1500 prepaces; extracellular ions also differ. Highest-dose APD90 median includes 484/500 valid cells and excludes unavailable crossings. Positive dose trend agrees, but magnitude is strongly overpredicted (396.3 ms versus 37 ± 7 ms). This does not validate quantitative APD, combined-block mitigation or clinical safety. Only IKr and ICaL potency are represented; neutral effects on other channels are assumptions.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U
AP prolongation and arrhythmic events
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
Barral 2025
Observed APD90 change (ms) · adult human ventricular trabeculae
APD90 prolongation
nominal; Table 2 says concentration was not bioanalytically measured
Source-derived observation records and per-study source notes in the audited original comparison
Barral 2025
Observed APD90 change (ms) · adult human ventricular trabeculae
APD90 prolongation
nominal; Table 2 says concentration was not bioanalytically measured
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
22. Ranolazine — Effect Predicted
The audited model output captures the reported dose-related ap prolongation.
Low-dose optical responses vary; cellular events are not equivalent to clinical TdP.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Dose-related AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
23. Risperidone — Effect Predicted
The audited model output captures the reported shift towards ap prolongation.
Both optical lines move towards prolongation; low-dose responses and other assay platforms vary.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Shift towards AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
24. Sotalol — Relevant Effect Captured
Prolongation captured; EADs were absent at all modeled active doses.
The APD direction agrees; the separately assessed EAD-like effect remains absent in tested simulations. EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
25. Tamoxifen — Effect Not Captured
The reported effect is not reproduced by the tested model outputs: APD90 prolongation.
iCell observations predominantly shorten; some other assay series prolong. The model prolongs throughout.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Shortening in iCell; assay-dependent response
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
26. Terfenadine — Relevant Effect Captured
Later prolongation and an EAD flag captured; initial shortening and high-dose quiescence are not fully reproduced.
Source low-dose shortening is missed; subsequent prolongation agrees. The highest source dose becomes quiescent.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
Initial shortening, then prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
27. Vandetanib — Effect Predicted
The audited model output captures the reported ap prolongation.
Main prolongation agrees; high-dose events limit eligible duration observations.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP prolongation
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
28. Verapamil — Effect Not Captured
The reported effect is not reproduced by the tested model outputs: APD90 prolongation.
The experimental shortening trend contrasts with increasing model APD90.
Blinova 2018
Observed ddAPD90c (ms) · Cor.4U; iCell2
AP shortening
Original Supplemental Table2 nominal bath concentrations, not free Cmax multiples. Base1uM is an exact unit conversion for importing the nominal concentrations as multipliers.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
29. Sertindole — Effect Predicted
The audited model output captures the reported published ead-like event.
EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles.
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
30. Erythromycin — Effect Not Captured
The reported effect is not reproduced by the tested model outputs: EAD not observed at tested doses.
EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles.
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
31. Sparfloxacin — Effect Not Captured
The reported effect is not reproduced by the tested model outputs: EAD not observed at tested doses.
EAD comparison uses presence at any tested concentration of the same molecule. Experimental one-hour observation differs from ten paced simulation cycles.
Huo 2019
EAD-like event presence · Preparation described in the linked primary article
Published EAD-like event
Source Table1 MEC-EAD: actual tested event-positive nominal concentration. Not IC50 or clinical free Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
32. Alfuzosin — Effect Predicted
The audited model output captures the reported fpd/apd60/qt prolongation.
FPD/APD60/QT and APD90 are different endpoints. The sodium-enhancement mechanism is absent from the inhibition-only model inputs.
Cardiac tissue 2022
Observed FPD change (%) · Vascularised human cardiac tissue
FPD/APD60/QT prolongation
Exactly0.3uM tested in primary human cardiac tissue FPD study and rabbit repolarization study. Do not substitute an invented Cmax.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
33. Lacosamide — Effect Predicted
The audited model output captures the reported apd90 shortening.
Rabbit APD90 means are compared with adult-human ORd; marginal SD does not establish uncertainty of paired changes.
Wolfes 2025
Observed APD90 change (ms) · Isolated rabbit whole-heart Langendorff, native perfusion, no sotalol
APD90 shortening
Exactly10and50uM in Wolfes2025 native rabbit-heart arm. Both finite reported concentrations; no Cmax required.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
34. BaCl2 — Effect Predicted
The audited model output captures the reported apd90 prolongation.
One measured concentration supports the direction; it does not establish a dose-response curve.
Jost 2013
Observed APD90 change (%) · Adult human papillary muscle
APD90 prolongation
10uM is the measured BaCl2 bath concentration in the source adult-human-tissue experiment. Base1uM is only an import unit conversion; no therapeutic Cmax is claimed.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
35. JNJ303 — Relevant Effect Captured
Positive APD80 point estimate captured; source p = 0.47 and FPD response depend on preparation.
APD80, not APD90, is plotted. The human-wedge increase is non-significant (p = 0.47); the 2 µM FPD response depends on cell line.
Kang 2017 / Europace
Observed APD80 change (ms) · Adult human LV wedge
Small or preparation-dependent repolarisation effect
Two independently published nominal bath concentrations. Observations retain separate assay and cell-system identities; no pooling or clinical-Cmax substitution.
Source-derived observation records and per-study source notes in the audited original comparison
Observed values by source and concentration
Model outputs by study and concentration
36. Ajmaline — Relevant Effect Captured
CARDIX reproduces native ventricular prolongation. Donor hiPSC cells showed slight APD90 shortening in another preparation; agreement is therefore qualified. Model EAD flags are not independently validated.
Native guinea-pig and donor/patient hiPSC preparations differ from the ORd population. Exposure and endpoint are not matched; suspected EAD is not confirmed TdP.
Effects of Ajmaline on Non-sodium Ionic Currents in Guinea Pig Ventricular Myocytes (Enomoto et al., 1995)
APD · Isolated guinea-pig ventricular myocytes
Dose-dependent APD prolongation was measured.
Concentration-dependent application; exact APD dose series not transcribed
Abstract final result; DOI 10.1536/ihj.36.465
A cellular model of Brugada syndrome with SCN10A variants using human-induced pluripotent stem cell-derived cardiomyocytes (2019)
APD50/APD90 · Healthy donor and Brugada patient hiPSC cardiomyocytes
Donor APD90 slightly shortened; Brugada APD slightly prolonged at 30 µM.
Sequential 3, 10 and 30 µM
Results: Differential effects of ajmaline; Figure 7
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
37. Amiodarone — Effect Predicted
Human ventricular APD90 prolongation agrees with the modeled direction at all active doses.
Loading/chronic treatment includes metabolites and remodeling absent from the acute 10-beat model. Oral mg/day is not converted to bath µM; conduction and amplitude were not comprehensively compared.
Frequency-dependent electrophysiologic effects of amiodarone in humans (Sager et al., 1993)
Ventricular APD90 · 19 loading and 15 chronic-treatment patients
APD90 increased about 30 ms, 10–13%, p < .001 at all paced cycles.
11 days loading at 1621 ± 162 mg/day; chronic 380 ± 56 mg/day; paced cycles 300–600 ms
Abstract methods/results; DOI 10.1161/01.cir.88.3.1063
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
38. Amitriptyline — Relevant Effect Captured
Small modeled shortening captures the initial dog response but not subsequent prolongation. A low-exposure canine wedge study found unchanged APD90; the small model response is compatible with that limited observation.
Model 0.0364–0.1456 µM is below the 0.2 µM wedge exposure and below toxic infusions. Infusion mg/kg/min cannot be treated as bath µM. The model does not represent tissue reentry.
Mechanisms of ventricular arrhythmia during amitriptyline toxicity
Monophasic APD50/APD90 · 23 anesthetized dogs
APD shortened then prolonged as serum concentrations rose; afterdepolarizations were not detected.
Graded infusion 0.5–1 mg/kg/min
Abstract results
Ionic and Cellular Mechanisms Underlying the Development of Acquired Brugada Syndrome in Patients Treated with Antidepressants
APD90 · Canine right-ventricular wedge
APD90 and QRS were unchanged; separate provocation conditions are not the monotherapy comparison.
0.2 µM amitriptyline alone
Results: Amitriptyline Induction of Phase 2 Reentry; Table 1A
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
39. Azithromycin — Effect Predicted
Modeled APD90 prolongation agrees with the positive QT association in the clinical ECG study. This supports effect direction rather than a clinical risk probability.
Clinical QTc is distinct from single-cell APD90. Observational associations can include disease and co-medication; not every clinical cohort shows prolongation. Independent guinea-pig ECG/APD evidence also supports prolongation, but at substantially higher exposures; it does not validate the modeled low-dose response.
Risk Evaluation of Azithromycin-Induced QT Prolongation in Real-World Practice (Choi et al., 2018)
QTc prolongation · 402,607 subjects, hospital ECG/EHR case-control analysis
QT prolongation OR 1.40, 95% CI 1.23–1.59; severe QT prolongation OR 1.43, 1.13–1.82.
Azithromycin prescription exposure; individual plasma concentrations unavailable
Abstract results; DOI 10.1155/2018/1574806
Electrophysiologic Studies on the Risks and Potential Mechanism Underlying the Proarrhythmic Nature of Azithromycin
QTc; APD50 and APD90 · Guinea-pig in-vivo ECG, isolated heart ECG and ventricular cardiomyocytes
Azithromycin prolonged measured QTc at high exposure; ventricular APD50/APD90 significantly lengthened at 830 mg/L.
114.6 mg/kg in vivo; 207.5/415 mg/L in isolated hearts; APD prolongation at 830 mg/L
Primary abstract Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
40. Chloroquine — Effect Predicted
Modeled prolongation agrees with measured feline Purkinje and ventricular APD prolongation over an overlapping exposure range.
Model 0.2495–0.998 µM overlaps only the lower part of the 0.3–10 µM experimental range. Higher-dose automaticity and resting-potential effects were not quantitatively validated.
Blockade of currents by the antimalarial drug chloroquine in feline ventricular myocytes (Sánchez-Chapula et al.)
APD · Cat Purkinje fibers and ventricular myocytes
APD increased with concentration; upstroke velocity fell, with automaticity and depolarization at higher doses.
0.3–10 µM
Abstract Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
41. Cibenzoline — Relevant Effect Captured
Modeled prolongation captures the canine ventricular branch. Published guinea-pig experiments report unchanged or shortened APD, so preparation-dependent findings are only partly reproduced.
Model 0.673–2.692 µM is mostly below the canine significant-effect exposure of 3 µM. Opposing guinea-pig findings remain explicit; the result is not a match across every preparation.
Effect of Cibenzoline, a Class I Antiarrhythmic Drug, on Action Potential in Canine Ventricular Muscle (Satoh et al., 1987)
APD90 · Canine right-ventricular papillary muscle, 60 beats/min
APD90 increased 26% at 3 µM and 30% at 8 µM, n = 6; no significant effect at 0.8 µM.
0.8, 3 and 8 µM; 15–20 minutes equilibration
Results page 114; Figure 1 and Table 1; DOI 10.1254/jjp.44.113
Effects of cibenzoline, a new class Ia antiarrhythmic drug, on various membrane ionic currents and action potentials of guinea-pig ventricular cells (Sato et al., 1994)
APD30/APD90 · Single guinea-pig ventricular cells, 0.2 Hz, 32–33°C
APD30 and APD90 shortened significantly at all tested exposures.
5, 10 and 30 µM
Abstract item 3; DOI 10.1007/BF00241092
Characterization of the class I antiarrhythmic activity of cibenzoline succinate in guinea pig papillary muscle (Arena et al., 1987)
APD50/APD90 · Guinea-pig papillary muscle
APD50 and APD90 were unchanged.
1–128 µM
Abstract
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
42. Cilostazol — Effect Predicted
Clinical QTc prolongation aligns with modeled APD90 prolongation. Raw QT did not change significantly in this cohort; the two endpoints are kept distinct.
Clinical disease, rate correction and chronic PDE/cAMP mechanisms differ from acute ORd channel scaling. Oral mg is not converted to µM. This is qualitative support, not clinical validation.
The Effect of Cilostazol on Electrocardiographic Parameters in Patients with Peripheral Artery Disease Initiating Cilostazol Treatment (Omar et al., 2023)
QTc · 32 patients with peripheral artery disease, before/after ECG
QTc increased from 414.34 ± 32.46 to 430.93 ± 33.25 ms, p = .01; raw QT was not significantly changed.
100 mg twice daily for 3 months
Abstract; Results; Table 2; DOI 10.51645/khj.2023.m347
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
43. Dasatinib — Relevant Effect Captured
The positive model sign agrees with clinical QTc prolongation, but every modeled APD change is below 1 ms and is displayed as Minimal APD90 change. Only a weak signal is captured.
Model changes of +0.07 to +0.37 ms are minimal under the existing descriptive rule. A sign match does not establish a clinically meaningful warning. The reported clinical 30 ± 9.7 ms change applies only to affected cases, not all 115 administrations.
Risk of QTc prolongation among cancer patients treated with tyrosine kinase inhibitors (Abu Rmilah et al., 2020)
QTcF · 115 dasatinib administrations in cancer patients
48/115 administrations showed QTc prolongation; selected affected cases had mean pre/post change 30 ± 9.7 ms.
20–140 mg/day among affected cases; most at least 70 mg/day
Table 2; Figure 1 affected cases; Table 3; DOI 10.1002/ijc.33119
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
44. Desipramine — Effect Predicted
The positive model APD trend agrees with primary acute ventricular observations. Chronic trafficking and apoptosis findings are limitations, not claimed model predictions.
Model 0.108–0.432 µM is below acute 1 and 10 µM experiments. APD100 differs from APD90; static block does not represent trafficking or apoptosis.
hERG K+ channel-associated cardiac effects of the antidepressant drug desipramine (Staudacher et al., 2011)
APD100 · Fresh guinea-pig ventricular myocytes; chronic cultures separately
Acute and chronic desipramine prolonged APD.
Acute 1 and 10 µM; overnight 30 µM
Abstract; Figure 8 in author PDF at https://publikationen.bibliothek.kit.edu/110082077/150185141; DOI 10.1007/s00210-010-0583-9
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
45. Diazepam — Effect Predicted
Every modeled absolute APD90 change is below 1 ms, agreeing with no APD prolongation from diazepam alone. Tiny modeled shortening is not presented as a measured clinical effect.
Model 0.029–0.116 µM is below the tested 1 µM. The less-than-1-ms rule is descriptive, not a statistical or safety threshold. The methadone interaction is a separate combination experiment.
Increased cardiac risk in concomitant methadone and diazepam treatment: pharmacodynamic interactions in cardiac ion channels
APD · Human stem cell-derived cardiomyocytes
Diazepam alone had no APD effect at 1 µM; combination-induced prolongation is a separate result.
Diazepam 1 µM alone; methadone combination studied separately
Abstract Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
46. Diphenhydramine — Effect Predicted
Small positive model APD changes agree with measured clinical and whole-heart repolarization lengthening.
Model 0.034–0.136 µM is below reported total plasma concentration around 0.7 µM and bath exposure of 10 µM. Free and total exposure are not interchangeable; clinical QT and MAPD are distinct from model APD90.
Block of potassium currents in guinea pig ventricular myocytes and lengthening of cardiac repolarization in man by the histamine H1 receptor antagonist diphenhydramine (Khalifa et al., 1999)
QTc/MAPD · Humans and isolated guinea-pig hearts
QTc increased above 20 ms in 7/20 patients; 10 µM lengthened MAPD; most usual-dose effects were modest.
Usual oral dose total plasma about 0.7 µM; heart bath 10 µM
Abstract clinical and isolated-heart results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
47. Donepezil — Effect Predicted
Modeled APD prolongation agrees with longer-term clinical QT/QTc increases. A separate 4-week study found no significant QT change; that negative result is retained as a qualification.
Not all cohorts show significant prolongation. Clinical QTc with disease and co-medication differs from single-cell APD90. Absence of model EAD flags does not establish absence of clinical TdP risk.
Long term use of donepezil and QTc prolongation
QT/QTc · 59 patients with before/after ECG
QT increased from 393.3 ± 35.6 to 411.9 ± 44.6 ms, p = .002; rate-corrected change also significant.
Long-term treatment; exact plasma µM unavailable
Abstract Results
Effect of QT Prolongation in Patients Taking Cholinesterase Inhibitors (Donepezil) for Alzheimer’s Disease
QTc · 57 treated patients and 57 controls; 46 before/after observations
Within 46 patients, QTc rose from 433 ± 34 to 442 ± 33 ms, p = .014.
Clinical donepezil treatment
Table 4
Electrocardiogram Changes of Donepezil Administration in Elderly Patients with Ischemic Heart Disease
QT/QTc · 60 patients with ischemic heart disease
No significant QT-parameter change; QTc 415.1 ± 35.7 to 424.3 ± 37.2 ms, p = .7529.
5 mg/day for at least 4 weeks
Abstract; Table 3
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
48. Duloxetine — Effect Predicted
Every modeled paired APD90 change is below 1 ms, compatible with absence of QT prolongation in the controlled clinical study. The observed small QTc decrease is not claimed to be reproduced.
Model +0.04 to +0.28 ms is displayed as Minimal APD90 change under the existing descriptive rule, not a statistical or safety threshold. This compares absence of prolongation; oral dose is not converted to µM.
QT effects of duloxetine at supratherapeutic doses: a placebo and positive controlled study (Zhang et al., 2007)
QTc · 117 healthy women; randomized placebo and moxifloxacin controlled crossover
No QT prolongation; mean QTc decreased versus placebo at 200 mg twice daily; no concentration relationship.
60 mg twice daily escalated to 200 mg twice daily; exposure above 5 times therapeutic
Abstract Results/Conclusion; DOI 10.1097/FJC.0b013e318030aff7
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
49. Flecainide — Effect Predicted
CARDIX prolongation is directionally consistent with measured rabbit QT and human stem-cell calcium-transient prolongation at 1 and 10 µM; these experimental endpoints differ from model APD90.
The 1uM source dose is approximately1.33x source free Cmax and falls within the model's0.752–3.008uM dose range;10uM is higher exposure.
Passini et al. 2017
QT interval · left ventricular rabbit wedge ECG
prolongation
Rabbit wedge ECG QT as described in Lu2016 and Passini2017; changes compared with baseline, vehicle QT cutoff5%.
Figure 5, Flecainide rows, experimental QT interval column only
Passini et al. 2017
CTD90 · spontaneously beating Cor.4U hiPSC-CMs, calcium fluorescence
prolongation
Cor.4U monolayers at37C; spontaneous calcium fluorescence using FDSS6000/Calcium5; CTD90 at25min after compound addition; source relevance cutoff25%.
Figure 5, Flecainide rows, experimental CTD90 column only
Observed values by source and concentration
Model outputs by study and concentration
50. Fluvoxamine — Effect Predicted
The positive model APD trend agrees with slight QTc prolongation in the guinea-pig experiment, without validating clinical magnitude or probability.
Systemic infusion is not converted to bath µM. QTc differs from APD90. Fluvoxamine produced a weaker signal than imipramine in the primary study; model +14 to +51 ms is not a validated effect size.
A Comparative Pharmacodynamic Study of the Arrhythmogenicity of Antidepressants, Fluvoxamine and Imipramine, in Guinea Pigs (Ohtani et al., 2001)
QTc · Anesthetized guinea-pigs with ECG and plasma concentration recording
Fluvoxamine prolonged QTc slightly; imipramine caused distinct dose-dependent prolongation.
Fluvoxamine 20 mg/kg/hour intravenously for 90 minutes
Abstract Results; DOI 10.1248/bpb.24.550
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
51. Halofantrine — Effect Predicted
Monotonic model APD90 prolongation agrees with the positive concentration–QTc relationship observed in humans.
Stereoselective free/total exposure, systemic dosing and malaria physiology differ from ORd channel block. Oral or intravenous mg is not converted to µM; no clinical risk probability is inferred.
Stereoselective halofantrine disposition and effect: concentration-related QTc prolongation
QTc · Healthy adults; 21 enrolled, 13 completed, 16 active and 5 placebo initially
14/16 active subjects had QTc prolongation positively correlated with both enantiomer concentrations.
Racemic halofantrine 500 mg/day for 42 days; serial enantiomer concentrations
Abstract Results/Conclusion
Pharmacokinetics, efficacy and toxicity of parenteral halofantrine in uncomplicated malaria (Krishna et al., 1993)
QT/QTc · 12 adults with acute malaria and 9 convalescent adults
After third dose, QTc prolonged 8.2 ± 5.6% (SD), p < .001.
1 mg/kg intravenously over 1 hour, every 8 hours, 3 doses
Abstract item 4; DOI 10.1111/j.1365-2125.1993.tb00419.x
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
52. Haloperidol — Relevant Effect Captured
The model captures the reported prolonging branch, but the primary experiments describe biphasic and preparation-dependent APD behavior. The selected model doses are also far below the cited 3 µM example. Only part of the reported response is captured, consistently with other partial compound reviews.
Model 0.004–0.016 µM is far below the 3 µM experimental example. Species and ischemia protocols differ. Only the prolonging branch is captured; high-dose biphasic behavior is not demonstrated.
Electrophysiological effects of haloperidol on isolated rabbit Purkinje fibers and guinea pigs papillary muscles under normal and simulated ischemia (Yan et al., 2007)
APD90 · Rabbit Purkinje fibers and guinea-pig papillary muscle
Under ischemia Purkinje APD90 prolonged concentration-dependently; papillary responses depended on condition.
Micromolar series; significant Purkinje prolongation at 3 µM
Abstract Results; DOI 10.1111/j.1745-7254.2007.00572.x
Comparison of guinea-pig ventricular myocytes and dog Purkinje fibres for in vitro assessment of drug-induced delayed repolarization (Terrar et al., 2007)
APD90 · Guinea-pig ventricular myocytes and dog Purkinje fibers
Haloperidol showed biphasic APD responses; the prolonging branch exceeded 25% in guinea-pig myocytes.
Full concentration grid not recovered
Primary PubMed abstract; DOI 10.1016/j.vascn.2007.04.005
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
53. Imipramine — Relevant Effect Captured
Model prolongation captures systemic QTc prolongation in Ohtani 2001. It does not reproduce shortening in bovine ventricular and canine Purkinje experiments; endpoint and preparation heterogeneity require a qualified partial result.
Model 0.106–0.424 µM is below the bovine shortening threshold above 1 µM. QTc includes conduction and rate correction and is not identical to cellular APD. No exposure-route conversion is attempted.
A Comparative Pharmacodynamic Study of the Arrhythmogenicity of Antidepressants, Fluvoxamine and Imipramine, in Guinea Pigs (Ohtani et al., 2001)
QTc · Anesthetized guinea-pigs
Distinct dose-dependent QTc prolongation.
Imipramine 10 and 20 mg/kg/hour intravenously for 90 minutes
Abstract Results; DOI 10.1248/bpb.24.550
Electrophysiological effects of imipramine on bovine ventricular muscle and Purkinje fibres
APD · Bovine ventricular muscle and Purkinje fibers
Ventricular APD shortened; Purkinje APD was unchanged; phase-2 amplitude decreased.
Ventricular shortening above 1 µM
Abstract items 3/4
Electrophysiologic effects of imipramine and doxepin on normal and depressed cardiac Purkinje fibers
APD/APA/Vmax · Canine Purkinje fibers
Dose-related decreases in APD, amplitude, upstroke velocity and conduction velocity.
Dose-related exposure; exact concentration grid unavailable
Abstract Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
54. Lidocaine — Relevant Effect Captured
CARDIX shortening agrees with measured rabbit QT shortening at 10 and 100 µM, while stem-cell CTD90 prolongs; the supporting literature therefore shows mixed endpoint agreement.
The10uM source dose is approximately3.85x Passini2017 free Cmax and lies near the model's highest dose10.2416uM;100uM is higher exposure. The selected model exposure is2.5604uM from Zhou2020 and remains distinct from Passini2017's2.6uM.
Passini et al. 2017
QT interval · left ventricular rabbit wedge ECG
shortening
Rabbit wedge ECG QT as described in Lu2016 and Passini2017; changes compared with baseline, vehicle QT cutoff5%.
Figure 5, Lidocaine rows, experimental QT interval column only
Passini et al. 2017
CTD90 · spontaneously beating Cor.4U hiPSC-CMs, calcium fluorescence
prolongation
Cor.4U monolayers at37C; spontaneous calcium fluorescence using FDSS6000/Calcium5; CTD90 at25min after compound addition; source relevance cutoff25%.
Figure 5, Lidocaine rows, experimental CTD90 column only
Observed values by source and concentration
Model outputs by study and concentration
55. Linezolid — Effect Predicted
The modeled shortening direction agrees with the directly measured early shortening after both clinical dose levels. Agreement is directional; the source effect was transient and much smaller than the modeled response, and the source does not establish sustained QT shortening.
Clinical QTcF and cellular APD90 are distinct endpoints. The experiment's oral/IV administered doses are not converted into fabricated µM coordinates. The short-lived observed effect is not evidence for a persistent clinical shortening of 15–21 ms.
Lack of an Effect of Standard and Supratherapeutic Doses of Linezolid on QTc Interval Prolongation
Time-matched placebo-corrected QTcF · Randomized placebo- and moxifloxacin-controlled four-way crossover in 40 healthy subjects
At 1 and 2 hours, QTcF changes were −2.96/−5.53 ms after 600 mg and −1.78/−7.51 ms after 1,200 mg. Shortening was absent by 4 hours; the study was negative for clinically significant QTc prolongation.
Single 600 or 1,200 mg IV infusions over 60 minutes
Results, Electrocardiography; Figure 1
Lack of an effect of standard and supratherapeutic doses of linezolid on QTc interval prolongation
QTcF · Healthy volunteers
Independent bibliographic verification of the same primary randomized trial, DOI 10.1128/AAC.01723-10.
600/1,200 mg IV; oral moxifloxacin control
Abstract and publication metadata
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
56. Lopinavir — Relevant Effect Captured
The model captures the direction of measured rabbit QT prolongation. It does not reproduce the same experiment's unchanged average APD90, the chronic hiPSC APD90 shortening, or observed acute EADs. These endpoint- and duration-dependent differences support partial agreement rather than full agreement.
The highest model dose is near, but below, the lowest rabbit dose of 3 µM; 5–10 µM is higher exposure. Chronic remodeling over 24 hours differs from the model's 10 paced cycles. QT, average APD90 and EAD presence must remain separate observations; no numerical EAD incidence is fabricated.
Electrophysiological Profile of Different Antiviral Therapies in a Rabbit Whole-Heart Model
QT, monophasic APD90 and repolarization dispersion · Langendorff-perfused AV-blocked rabbit hearts; lopinavir group n=12
QT increased from 253 ms to 257, 268 and 278 ms, whereas average APD90 was not significantly changed (168 ms baseline; 163, 168, 164 ms).
3, 5 and 10 µM isolated lopinavir, without ritonavir; paced cycle lengths 300–900 ms
Results, Lopinavir; Figure 4
Chronic Administration of COVID-19 Drugs Fluvoxamine and Lopinavir Shortens Action Potential Duration by Inhibiting the Human Ether-à-go-go–Related Gene and Cav1.2
APD90 and EAD presence · hiPSC-derived ventricular-like cardiomyocytes
APD90 fell from 364.4 to 172.6 ms after 24 hours (n=14). EADs were recorded during acute treatment.
10 µM lopinavir; 24-hour incubation for duration, within 1 hour for EAD examples
Results, Figure 6C–H and associated text
Observed values by source and concentration
Model outputs by study and concentration
57. Methadone — Effect Predicted
Modeled concentration-dependent prolongation agrees with measured QTc prolongation in a randomized methadone trial. This verdict concerns repolarization direction, not prediction of individual clinical QTc magnitude or TdP.
Clinical QTc is not cellular APD90. Long-term clinical treatment, pharmacokinetics and patient comorbidity differ from 10-cycle cellular simulation. No clinical EAD or TdP claim is inferred from the absence of a cellular EAD flag.
QT-interval effects of methadone, levomethadyl, and buprenorphine in a randomized trial
QTc and increase from baseline · 17-week randomized double-blind trial; ECGs from 165 opioid-dependent participants
Methadone-treated participants developed prolonged QTc and increases from baseline more frequently than buprenorphine-treated participants.
Methadone maintenance treatment with ECGs at baseline and every 4 weeks
Abstract, Results; JAMA full text Figure 2
QT-Interval Effects of Methadone, Levomethadyl, and Buprenorphine in a Randomized Trial
Longitudinal QTc · Human randomized trial
The original primary trial demonstrates an increasing QTc trajectory under methadone.
Maintenance treatment over 17 weeks
Results and Figure 2
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
58. Metronidazole — Effect Predicted
The model produces the reported prolongation direction at higher tested concentrations. Under the stated any-modeled-dose criterion this captures the observed effect; it does not establish the source patient's exposure, the mechanism, or the clinical magnitude.
Single susceptible patient; concomitant medications and an unproven mechanism limit causal generalization. No µM concentration is recoverable for the clinical observation, so no false dose-response point is plotted. The modeled initial shortening has no corresponding observation in this case report.
QT interval prolongation due to metronidazole administration
Clinical QTc · 71-year-old woman with multiple cardiovascular/pulmonary conditions and prior drug-induced QT prolongation
Primary case report of QTc prolongation during metronidazole treatment.
IV metronidazole; change occurred after 2 days; exact administered dose and plasma concentration not stated in the recovered report
Primary letter metadata, DOI 10.5152/akd.2011.120
QT interval prolongation due to metronidazole administration
QTc before and after treatment · Clinical case report
QTc increased from 396 to 559 ms after 2 days, with normal potassium/magnesium; ECG normalized within 48 hours after discontinuation.
IV metronidazole for nosocomial pneumonia
Page 468, case paragraph and Figures 1–2; page 469, susceptibility and mechanism discussion
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
59. Mibefradil — Effect Not Captured
The selected model regimen does not reproduce the primary study's consistent cellular shortening or in-vivo absence of QT prolongation. A nonsignificant slight lengthening in a different rabbit high-exposure preparation is retained as mixed context, not treated as confirmed prolongation matching the modeled trend.
The source comprises multiple species, endpoints and exposure levels. The complete cellular concentration grid was not recovered; this is a direction comparison, not a dose-matched challenge. No statistical confidence is assigned to the source's nonsignificant rabbit lengthening.
Effects of the T-type Ca(2+) channel blocker mibefradil on repolarization of guinea pig, rabbit, dog, monkey, and human cardiac tissue
Cardiac AP/MAP duration and QT · Isolated guinea-pig and human cardiomyocytes, rabbit hearts, open-chest bradycardic dogs and other preparations
Mibefradil shortened human/guinea-pig AP plateau; dog MAPs shortened slightly with no QT effect. Rabbit high cardiodepressant concentrations produced slight nonsignificant lengthening.
Dose-dependent cellular effects and noncardiodepressant/cardiodepressant whole-heart exposure; exact concentration range not recovered from abstract
Primary abstract, Results and Conclusion
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
60. Mitoxantrone — Effect Not Captured
Neither the reported prolongation direction nor EAD presence was observed anywhere in the selected modeled range. The primary experimental dose of 30 µM is approximately 33 times the highest modeled dose, so the result is a failure to capture this literature effect in the chosen regimen, not evidence that CARDIX would fail at 30 µM.
The relevant experimental effect was measured at much higher concentration than the completed model grid. Species and time dependence differ from the 10-cycle human ORd simulation. No numerical APD or EAD incidence is invented from the qualitative abstract.
Effects of mitoxantrone on action potential and membrane currents in isolated cardiac myocytes
Ventricular APD and EAD presence · Whole-cell-clamped isolated guinea-pig ventricular myocytes
Mitoxantrone prolonged APD and sometimes induced EADs; IKr and IK1 depression was measured.
30 µM mitoxantrone; time-dependent effect and reverse rate dependence
Abstract items 2–3 and conclusion; PMID 10385229
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
61. Moxifloxacin — Effect Predicted
CARDIX prolongation is directionally consistent with measured rabbit QT increases at 10–30 µM and stem-cell CTD90 prolongation at 30 µM.
The10/30uM source doses are approximately0.91/2.74x source free Cmax, bracketing or within the model's10.96–43.84uM range;300uM is higher exposure. QT at300uM and CTD90 at10uM were not tested.
Passini et al. 2017
QT interval · left ventricular rabbit wedge ECG
prolongation
Rabbit wedge ECG QT as described in Lu2016 and Passini2017; changes compared with baseline, vehicle QT cutoff5%.
Figure 5, Moxifloxacin rows, experimental QT interval column only
Passini et al. 2017
CTD90 · spontaneously beating Cor.4U hiPSC-CMs, calcium fluorescence
prolongation
Cor.4U monolayers at37C; spontaneous calcium fluorescence using FDSS6000/Calcium5; CTD90 at25min after compound addition; source relevance cutoff25%.
Figure 5, Moxifloxacin rows, experimental CTD90 column only
Observed values by source and concentration
Model outputs by study and concentration
62. Nilotinib — Effect Predicted
The model's increasing prolongation across concentration agrees with the measured clinical exposure-QTcF direction. This does not imply numerical QTcF calibration, individual clinical arrhythmia prediction, or equal free/total concentrations.
Clinical QTcF is not cellular APD90; the clinical data do not validate the modeled magnitude. Serum total concentration is distinct from the model's free concentration grid. The overall clinical cohort median was stable despite individual prolongation; the positive exposure-response study supplies the graded direction.
Population pharmacokinetic and exposure-response analysis of nilotinib in patients with newly diagnosed Ph+ chronic myeloid leukemia in chronic phase
Time-matched serum nilotinib concentration and QTcF change · Clinical ENESTnd exposure-response analysis; 542 patients contributed pharmacokinetic data
A positive correlation was observed between measured nilotinib concentration and QTcF change from baseline.
Nilotinib 300 or 400 mg twice daily; concentration/QTcF pairs on days 8 and 84
Results, Exposure-safety relationship; Figure 5
Clinical cardiac safety profile of nilotinib
QTcF longitudinal follow-up · 81 nilotinib-treated CML patients
QTcF increases greater than 30 ms occurred in 18 patients and greater than 60 ms in two; cohort median QTcF did not change significantly.
Median treatment 26 months
Results, Cardiac function and Figure 1A
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
63. Nimodipine — Effect Predicted
CARDIX shortening shares the measured rabbit QT direction at 0.1–1 µM; these experiments used much higher concentrations than this study's 1–4× exposure range.
Only higher-exposure direction support:0.1/1uM are100/1000x source free Cmax, above the model's0.001–0.004uM range. CTD90 changes(+7/−14%) do not exceed the source's25% assay relevance cutoff.
Passini et al. 2017
QT interval · left ventricular rabbit wedge ECG
shortening
Rabbit wedge ECG QT as described in Lu2016 and Passini2017; changes compared with baseline, vehicle QT cutoff5%.
Figure 5, Nimodipine rows, experimental QT interval column only
Passini et al. 2017
CTD90 · spontaneously beating Cor.4U hiPSC-CMs, calcium fluorescence
small mixed changes below source relevance cutoff
Cor.4U monolayers at37C; spontaneous calcium fluorescence using FDSS6000/Calcium5; CTD90 at25min after compound addition; source relevance cutoff25%.
Figure 5, Nimodipine rows, experimental CTD90 column only
Observed values by source and concentration
Model outputs by study and concentration
64. Nisoldipine — Effect Predicted
The model captures the repeatedly measured shortening direction. The recovered experiments used much higher concentrations and some diseased-cell preparations, so this is qualitative direction support rather than validation of the response at the model's very low exposure.
The lowest recovered experimental concentration, 0.05 µM, is 125 times the highest model concentration; 0.2–1 µM is still higher. The human cell experiment used a familial LQT2 phenotype; it is not a healthy-cell quantitative benchmark. Published difference-of-means is labeled as such, not paired-change SEM.
Multiple modulations of action potential duration by different calcium channel blocking agents in guinea pig ventricular myocytes
APD90 · Whole-cell patch-clamped guinea-pig ventricular myocytes
Nisoldipine shortened APD90 without a lengthening phase.
0.2–1 µM nisoldipine
Abstract, Results
The development of L-type Ca2+ current mediated alternans does not depend on the restitution slope in canine ventricular myocardium
Baseline ventricular APD90 · Canine ventricular myocardium; 500-ms baseline cycle length
APD90 decreased from 187±5 to 159±5 ms, p<0.05, n=7.
1 µM nisoldipine; n=7
Results, ICaL inhibition suppresses alternans; Figure 6
Physiological genomics identifies genetic modifiers of long QT syndrome type 2 severity
APD90 · Patient-derived human iPSC-CMs with severe familial LQT2
Nisoldipine shortened the prolonged APD90 in severely affected patient-derived cells toward the family-control value.
0.05 µM nisoldipine
Figure 5 and Supplemental Table 4
Observed values by source and concentration
Model outputs by study and concentration
65. Paliperidone — Effect Predicted
Modeled prolongation agrees with directly measured QTc prolongation reported in the primary regulatory trial summary. The verdict is directional; the clinical QT effect was modest and does not calibrate the modeled APD90 magnitude.
The positive 8-mg observation is from an immediate-release formulation, distinct from the extended-release comparator trial. No oral dose or single Cmax summary is fabricated into a model-matched free µM dose grid. Clinical QTcLD and cellular APD90 are different endpoints.
INVEGA HAFYERA data sheet: Effect on QT/QTc interval and cardiac electrophysiology
Placebo-subtracted QTcLD · Double-blind active-controlled thorough-QT study in adults with schizophrenia/schizoaffective disorder; total n=141
Mean placebo-subtracted QTcLD increase was 12.3 ms (90% CI 8.9–15.6) for the 8-mg immediate-release dose.
Immediate-release oral paliperidone 8 mg; n=50; day 8 at 1.5 hours post-dose; Cmax 113 ng/mL
Clinical pharmacology, Effect on QT/QTc interval and cardiac electrophysiology
Evaluation of the effect of paliperidone extended release and quetiapine on corrected QT intervals: a randomized, double-blind, placebo-controlled study
QTcLD · Randomized 109-patient clinical trial
Primary additional QT study verifies testing at therapeutic and supratherapeutic paliperidone ER doses; it is an active-comparator noninferiority assessment rather than the source of the 12.3-ms immediate-release result.
12 and 18 mg/day paliperidone ER versus quetiapine 800 mg/day and placebo
Abstract, primary and secondary QT comparisons
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
66. Paroxetine — Relevant Effect Captured
The small positive modeled response is compatible in direction with rare reported QTc increases, but does not capture the explicitly measured cellular APD90 shortening. Controlled clinical evidence was predominantly near-null; therefore this is mixed evidence and partial agreement, not a claim that paroxetine generally prolongs clinical QT or that a new no-effect threshold has been met.
The observed 0.4-µM cellular shortening was at more than seven times the highest modeled concentration. The rare clinical QTc cases do not establish a general population trend, magnitude or a matched concentration. The clinical near-null result is not converted into confirmed positive prolongation or a retrospectively chosen success threshold. Peak voltage alone is not AP amplitude or maximum upstroke velocity; those endpoints are not interchanged.
[ECG changes after paroxetine. 3 case reports]
Clinical QTc · Three high-risk clinical case reports
QTc increased in two reported cases; the patients had a high-risk clinical profile.
Paroxetine treatment; quantitative exposure was not recoverable from abstract
Primary English abstract
Effects of Paroxetine on a Human Ether-a-go-go-related Gene (hERG) K+ Channel Expressed in Xenopus Oocytes and on Cardiac Action Potential
APD90 · Guinea-pig ventricular myocytes at 36°C
The primary research abstract explicitly reports an APD90 decrease of 4.3%; hERG block alone is not substituted for this measured direction.
0.4 µM paroxetine for 5 minutes
Primary article abstract; International Journal of Oral Biology 2018, 43(1):43–51, DOI 10.11620/ijob.2018.43.1.043
An open-label, single-arm, dose-escalating concentration-QT study to investigate the cardiac effects and safety of paroxetine in healthy adults
QTcF versus plasma concentration · 38 healthy adults
The concentration-QTc analysis found no clinically significant effect; estimated maximum change was +0.42 ms with a 90% CI crossing zero at 60 mg/day.
20, 40 and 60 mg/day, one week per dose level
Abstract, Results and Conclusions
The Antidepressant Paroxetine Reduces the Cardiac Sodium Current
APD90, AP amplitude and maximum upstroke velocity · Six isolated rabbit left-ventricular cardiomyocytes
AP amplitude and upstroke velocity decreased, while APD20/50/90 did not change significantly.
3 µM paroxetine; paced at 1 Hz
Results section 2.2; Figure 4
Observed values by source and concentration
Model outputs by study and concentration
67. Pentobarbital — Effect Predicted
The positive APD direction agrees with measured ventricular prolongation. The model response is very small and the published anesthetic exposures were higher; only direction is supported, not the experimental magnitude or an adequate clinical liability warning.
APD95 is distinct from modeled APD90. Published concentrations are mg/L and have not been converted into µM without confirming the chemical-form basis. The modeled effect is orders smaller than the primary experiment's percentage changes; this verdict is qualitative only.
Direct electrophysiological actions of pentobarbital at concentrations achieved during general anesthesia
APD95, effective refractory period and maximum upstroke velocity · Guinea-pig papillary muscle and canine ventricular muscle
Pentobarbital increased guinea-pig APD95 by 24±6% and 33±4% at 25/50 mg/L; dog ventricular APD also increased concentration-dependently.
Guinea pig: 25 and 50 mg/L; dog ventricular effects studied over 5–100 mg/L
Primary abstract, Results; DOI 10.1152/ajpheart.1990.259.6.H1743
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
68. Phenytoin — Effect Predicted
The modeled shortening direction is consistent with experimentally observed action-potential plateau shortening.
Published measurements describe plateau shortening in isolated calf/dog Purkinje fibers. CARDIX reports APD90 in adult human ventricular ORd cells. This supports qualitative direction without establishing equal APD90 magnitude or dose response. The modeled range overlaps the lower part of the 5–100 µM experimental range. The lowest model dose, 4.36 µM, is below the published range; 17.44 µM is the model maximum, not the published maximum of 100 µM.
Scheuer and Kass 1983
Action-potential plateau phase and twitch tension · Isolated calf and dog cardiac Purkinje fibers
Lower/shorter plateau and reduced twitch tension
Electrical/mechanical measurements; two-microelectrode voltage clamp for ionic currents. Pacing, temperature and n were not specified in the abstract.
Abstract
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
69. Prenylamine — Effect Predicted
The model captures the measured clinical prolongation direction. The primary comparison concerns QT duration, not a claim that the model has reproduced prenylamine-associated TdP.
Clinical QT is distinct from cellular APD90. Oral treatment over weeks/months and free plasma concentrations are not the same as the 10-cycle model grid. The primary study did not observe serious arrhythmic events, so absent simulated EADs are not marked as a mismatch for this specific comparison.
The effect of prenylamine on the QT interval of the resting electrocardiogram in patients with angina pectoris
Resting ECG QT interval · 29 patients with angina pectoris followed before, during and after treatment
QT became significantly prolonged after one week, remained prolonged during treatment and returned to normal within two weeks of withdrawal.
Prenylamine 180 mg daily for up to 6 months
Primary abstract and Results; DOI 10.1136/pgmj.56.661.753
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
70. Primidone — Effect Not Captured
The model does not produce the reported shortening direction at any selected concentration. The source involved congenital long-QT patients and long-term treatment, whereas the simulation used a normal ventricular population and direct parent-drug channel scaling; this limits generalization of the mismatch.
Small case series in congenital long-QT syndrome, not a controlled healthy-cell experiment. Long-term treatment and active metabolite contributions are not reproduced by a 10-cycle parent-drug ORd simulation. No clinical dose is converted into a fabricated µM graph point.
Primidone in the treatment of the long QT syndrome: QT shortening and ventricular arrhythmia suppression
QT interval and ventricular arrhythmia · Three members of a family with long-QT syndrome
Electrocardiographic and Holter recordings documented QT shortening during primidone treatment; ventricular arrhythmias were suppressed in the index case.
Primidone therapy; index-case follow-up 2 years and two relatives followed for 8 months; quantitative dose/plasma concentration not recovered from abstract
Primary abstract; DOI 10.7326/0003-4819-93-1-53
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
71. Procainamide — Effect Predicted
The model's prolongation direction agrees with measured QTc and late-repolarization APD changes. Effects on earlier repolarization and depolarization vary by preparation; the verdict does not claim those other endpoints were uniformly predicted.
Clinical QTc, late APD100 and modeled APD90 are distinct endpoints. A reported serum concentration interval is not collapsed into an invented single experimental µM coordinate. The clinical parent drug/metabolite mix and protein binding differ from direct fixed-exposure cellular simulation.
Effects of quinidine versus procainamide on the QT interval
QTc change from baseline · 18 patients receiving quinidine and procainamide separately
Procainamide increased QTc by a reported mean 39±7 ms (SEM); its effect was smaller than quinidine's.
Procainamide titrated to trough serum concentrations 4–12 µg/mL
Primary abstract; DOI 10.1016/0002-9149(86)90025-1
Effects of therapeutic concentrations of procainamide on transmembrane action potentials of normal and infarct zone Purkinje fibers and ventricular muscle cells
Late APD100 and depolarization parameters · Isolated canine Purkinje fibers and left-ventricular endocardial muscle
Procainamide increased APD100 in normal ventricular muscle and altered Purkinje APD in a preparation-dependent manner.
10 mg/L procainamide
Abstract, Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
72. Propafenone — Effect Predicted
The modeled ventricular duration and peak-voltage directions agree with measured isolated ventricular-muscle effects. Purkinje-fiber shortening in the same article is a different cell preparation and is not represented by this ventricular-cell family.
Direction comparison is restricted to the ventricular-muscle preparation. Purkinje behavior, conduction velocity and re-entry are outside the single-cell model. No numerical APD90 concentration-response values were recovered from the accessible primary abstract.
Propafenone slows conduction and produces a nonuniform recovery of excitability between Purkinje and ventricular muscle fibers
APD90, action-potential amplitude, Vmax · Isolated sheep ventricular muscle and Purkinje fibers
At 0.1–1 µM the drug reduced action-potential amplitude and Vmax; ventricular APD90 lengthened while Purkinje APD90 shortened.
0.1–1 µM; higher concentrations also examined
Primary article abstract; DOI 10.1097/00005344-199308000-00005
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
73. Propranolol — Effect Not Captured
The model's duration trend is opposite to measured concentration-dependent APD shortening in human papillary muscle and MAP90 shortening in patients. A modest model peak decrease does not reproduce the principal duration effect.
Clinical MAP and isolated papillary-muscle APD are distinct preparations from an ORd cell. Clinical dose and tissue mass-concentration units are retained; no unsupported conversion to simulated free µM concentrations.
Electrophysiologic effects of propranolol on the human heart
Action-potential duration, dV/dt, amplitude · Human papillary muscle collected during cardiac surgery
Propranolol shortened APD concentration-dependently and lowered depolarization dV/dt; very high concentration also lowered amplitude.
10^-9–10^-4 g/mL; APD shortening reported at 10^-8–10^-6 g/mL
Primary article abstract; DOI 10.1536/ihj.19.136
Effects of propranolol on ventricular repolarization in man
MAP90 at constant paced heart rate · Ten patients with coronary artery disease
Median MAP90 shortened from 238 to 228 ms while ventricular effective refractory period remained unchanged.
0.2 mg/kg intravenous; pacing 100 beats/min
Primary article abstract; DOI 10.1007/BF00280044
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
74. Quetiapine — Effect Predicted
The positive modeled duration trend agrees qualitatively with measured clinical QTc prolongation. This is directional agreement, not agreement in effect magnitude or a same-endpoint validation.
QTcI is a whole-heart clinical endpoint; CARDIX reports single-cell APD90. Oral 100 mg was not converted to µM. No numerical experimental points are placed on the model concentration axis.
A thorough QT study to evaluate the QTc prolongation potential of two neuropsychiatric drugs, quetiapine and escitalopram, in healthy volunteers
Placebo-adjusted change in QTcI · Forty healthy volunteers; randomized four-period crossover
Quetiapine produced a maximum mean QTcI increase of 10.2 ms, with an upper 95% confidence bound of 13.7 ms.
Single oral quetiapine immediate-release 100 mg
Abstract Results; DOI 10.1097/YIC.0000000000000124
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
75. Quinine — Effect Predicted
The direction of model prolongation agrees with the positive measured QT response. The magnitude and exposure are not matched, and a separate antiarrhythmic patient study reported no QTc prolongation, illustrating preparation and protocol dependence.
The selected positive observation concerns QT/QaTc, not human-cell APD90; the perfused-heart concentration range is below the simulated range. Separate human trial results were not uniformly positive. Agreement means reproduction of a documented prolonging direction, not consistency across all protocols. No numerical ECG dose-response values were digitized from figures.
Effects of anti-malarial drugs on the electrocardiographic QT interval modelled in the isolated perfused guinea pig heart system
QaT / corrected QaTc duration · Isolated perfused guinea-pig hearts paced at 210 beats/min
Quinine lengthened the measured ECG repolarization interval in the concentration-response assay.
Quinine 0.3–2.4 µM
Methods QT interval measurement and measured QT-prolongation results; DOI 10.1186/1475-2875-9-318
Quinine sulfate prescribing information: controlled cardiac electrophysiology study
Baseline- and placebo-adjusted QTcI · Young and elderly healthy subjects, 13 in each age group
Maximum mean QTcI increase was 27.7 ms, with an upper 95% confidence bound of 32.2 ms.
Quinine sulfate 648 mg three times daily for seven days
Section 12.2 Pharmacodynamics
Antiarrhythmic activity of quinine in humans
QTc and ventricular refractoriness · Patients with ventricular ectopy / inducible ventricular tachycardia
The report found antiarrhythmic activity and refractory-period prolongation but did not find QTc prolongation.
Effective serum concentration mean 11 µM, range 4–17 µM in the first trial
Primary article abstract Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
76. Raltegravir — Effect Predicted
The nearly unchanged model duration and absence of suspected EADs agree with the negative thorough-QT finding. The <1 ms description is an internal descriptive rule, not a statistical or clinical threshold.
Measured clinical total plasma Cmax is not identical to a protein-free model concentration. A negative QTcF result and a <1 ms model APD90 change are a qualitative comparison, not interchangeable quantitative endpoints.
Raltegravir thorough QT/QTc study: a single supratherapeutic dose of raltegravir does not prolong the QTcF interval
Placebo-adjusted QTcF change · Healthy volunteers; randomized double-blind three-period crossover
Raltegravir did not prolong QTcF; the upper two-sided 90% confidence bound stayed below 10 ms at every assessed time.
Single 1600 mg oral dose; measured mean total Cmax approximately 20 µM
Primary article abstract; DOI 10.1177/0091270008318007
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
77. Ritonavir — Relevant Effect Captured
The positive duration direction captures the modest high-dose clinical QTc signal, but not the negative 100 mg QT study or the opposite APD shortening measured in isolated rabbit cells. The result is mixed supporting evidence.
Observed APD shortening is at a higher bath concentration than the model and involves a chloride-current mechanism outside standard ORd. QTcF and isolated APD90 are different endpoints; the clinical positive signal is modest and does not imply a positive regulatory thorough-QT result. The quantitative rabbit point must remain a separate endpoint/preparation series and must not be used as clinical QTc.
Ritonavir prescribing information: cardiac electrophysiology
Placebo- and baseline-adjusted QTcF / PR changes · Forty-five healthy adults; randomized active- and placebo-controlled crossover
The maximum mean QTcF increase was 5.5 ms, with an upper 95% confidence bound of 7.6 ms; PR also increased.
400 mg twice daily, assessed on day 3
Section 12.2 Pharmacodynamics
Ritonavir 100 mg does not cause QTc prolongation in healthy subjects: a possible role as CYP3A inhibitor in thorough QTc studies
Placebo-adjusted QTcF change · Sixty-five healthy volunteers; randomized crossover
Maximum mean placebo-adjusted QTcF increase was 0.16 ms, with a 90% confidence interval from −1.38 to 1.69 ms; the study was negative.
Single oral 100 mg
Primary article abstract; DOI 10.1038/sj.clpt.6100263
HIV protease inhibitors elicit volume-sensitive chloride current in cardiac myocytes via mitochondrial ROS
APD50, APD90 · Isolated rabbit ventricular myocytes
APD90 decreased by 60 ± 9 ms in four cells; the change was reversed by a volume-sensitive chloride-current blocker.
Ritonavir 15 µM for 15 minutes
Section 3.5 and Figure 6; DOI 10.1016/j.yjmcc.2010.08.013
Observed values by source and concentration
Model outputs by study and concentration
78. Rufinamide — Effect Predicted
The modeled repolarization-shortening direction is consistent with the published clinical QTc-shortening effect.
Clinical QTc and single-cell APD90 are different endpoints. Directional consistency does not establish numerical equality, concentration equivalence or clinical safety. The primary abstract supplies no comparable free plasma or bath-concentration grid. No conversion from oral mg/day to modeled µM is assumed.
Schimpf et al. 2012
Clinical QT/QTc intervals before and during oral therapy · 19 patients with difficult-to-treat epilepsies
QTc shortening
Before/during oral treatment. Figure 2 shows one patient at 2400 mg/day with concomitant antiepileptics; this is not an assay bath concentration.
Abstract; Figure 2 caption
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
79. Saquinavir — Relevant Effect Captured
The modeled prolongation direction agrees with the measured saquinavir-containing regimen. The experiment used ritonavir boosting, so attribution and exposure cannot be validated as saquinavir monotherapy.
The observed regimen includes ritonavir, whereas this CARDIX study is saquinavir alone. The source's delayed/nonlinear clinical exposure-response is not represented by an acute ten-cycle fixed-concentration model. Oral dose was not mapped to simulated free µM concentrations.
Thorough QT/QTc study of ritonavir-boosted saquinavir following multiple-dose administration of therapeutic and supratherapeutic doses in healthy participants
Baseline- and placebo-adjusted study-specific QTc · Healthy participants; randomized double-blind four-way crossover
The greatest mean QTc increases were 18.9 and 30.2 ms for the therapeutic and supratherapeutic regimens.
Saquinavir/ritonavir 1000/100 mg and 1500/100 mg twice daily
Primary article abstract; DOI 10.1177/0091270011400071
FDA saquinavir clinical review: QT and PR study NP21249
QTc and PR · Healthy-volunteer saquinavir/ritonavir study
The primary trial showed dose-dependent QTc and PR prolongation; QTc exposure-response was not linear and showed a delayed effect.
1000/100 mg and 1500/100 mg twice daily
Section 6 Assessment of Safety, QT-study findings
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
80. Sitagliptin — Effect Predicted
The model's minimal change agrees qualitatively with the controlled study's absence of clinically meaningful QT prolongation and shallow positive exposure-response.
A <1 ms model APD90 description is not the clinical QTc decision threshold. The reported slope is a clinical pharmacokinetic-regression result, not an observed APD90 concentration point; it was not plotted as experimental samples.
A thorough QTc study to assess the effect of sitagliptin, a DPP4 inhibitor, on ventricular repolarization in healthy subjects
Placebo-adjusted QTcF change and concentration-QTc relationship · Healthy volunteers; randomized double-blind four-period crossover
The clinical dose did not increase QTcF; the supratherapeutic dose produced only clinically insignificant prolongation, with a modeled slope of 0.59 ms per 1000 nM plasma concentration.
Single oral 100 mg and 800 mg; 800 mg yielded approximately eleven-fold higher maximal concentrations
Primary article abstract; DOI 10.1177/0091270009337511
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
81. Solifenacin — Effect Predicted
The positive model duration trend agrees with the small dose-related measured QTcF response. This comparison is qualitative and does not equate oral dosing with model concentration.
QTcF is a clinical endpoint, not the same measurement as model APD90. No numerical clinical points were assigned to unverified free µM concentrations.
VESIcare prescribing information: cardiac electrophysiology trial
Placebo-adjusted QTcF at Tmax · Seventy-six healthy women; multidose randomized active- and placebo-controlled study
Mean QTcF changes were 2 ms (90% CI −3 to 6) and 8 ms (90% CI 4 to 13) at 10 and 30 mg; higher-dose prolongation exceeded the lower-dose effect.
Solifenacin 10 mg and 30 mg
Cardiac Electrophysiology and Table 1, printed page 4
Solifenacin succinate prescribing information: QT prolongation
QTcF · Seventy-six healthy women
The controlled mean QTcF increase was 8 ms with 90% CI 4–13, and the prolonging effect was smaller at 10 mg.
30 mg, three times the maximum adult dose
Sections 5.6 and 12.2
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
82. Sunitinib — Effect Predicted
Modeled duration increases with concentration, agreeing with the measured exposure-related QTcF direction. The model does not reproduce cancer-patient physiology, metabolite contributions or clinical effect magnitude.
Clinical measurements include active-metabolite exposure and patient covariates; this CARDIX study uses parent-drug fixed free concentrations. Oral loading/maintenance regimens were not converted to model µM points.
Electrocardiographic characterization of the QTc interval in patients with advanced solid tumors: pharmacokinetic-pharmacodynamic evaluation of sunitinib
Time-matched placebo-adjusted QTcF · Cancer patients; 24 of 48 were QT/PK evaluable
QTcF changes increased with exposure. Maximum mean changes were 9.6 ms (90% CI 4.1–15.1) at therapeutic exposure and 15.4 ms (8.4–22.4) at supratherapeutic exposure.
Loading 150–200 mg on days 3 and 9; maintenance 50 mg/day on days 4–8
Primary article abstract Results; DOI 10.1158/1078-0432.CCR-09-1521
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
83. Tedisamil — Effect Predicted
The modeled duration direction agrees with measured human ventricular repolarization prolongation and isolated ventricular-muscle observations.
The clinical dose and reported isolated-fiber dose do not match the exact simulated concentration grid. This verdict addresses duration direction; it does not claim the model reproduces human TdP or all rate-dependent mechanisms.
Prolongation of monophasic action potential duration and the refractory period in the human heart by tedisamil, a new potassium-blocking agent
MAP90, QTc · Ten patients with coronary artery disease; atrial pacing at multiple cycle lengths
QTc lengthened by about 10% and ventricular MAP90 by about 16%; prolongation persisted during constant-rate pacing.
0.3 mg/kg intravenous
Primary article abstract; DOI 10.1093/oxfordjournals.eurheartj.a060403
Frequency-dependent cardiac electrophysiologic effects of tedisamil: comparison with quinidine and sotalol
AP duration · Isolated dog ventricular muscle / Purkinje fibers and rabbit atrial muscle
Tedisamil lengthened repolarization in ventricular muscle and Purkinje fibers, with preparation-dependent frequency effects.
Tedisamil 1 µM in reported Purkinje comparison
Primary article abstract Methods and Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
84. Telbivudine — Effect Not Captured
The monotonic model duration increase does not reproduce the controlled study's lack of QTcF prolongation or concentration-related trend. APD90 and QTcF are different endpoints, so the mismatch is qualitative rather than a clinical-threshold comparison.
No clinical mean per-dose values or matched free-plasma µM observations were recovered; no artificial zero-valued points were plotted. The discrepancy may reflect the supplied ion-channel input profile and does not establish clinical proarrhythmia.
FDA telbivudine clinical pharmacology review
Thorough QT / QTcF · Healthy volunteers; placebo- and positive-controlled study
Both clinical and supratherapeutic doses met the criterion for a negative thorough-QT study; moxifloxacin was positive.
Telbivudine 600 mg/day and 1800 mg/day
Section 2.2.4.3, Does this drug prolong the QT or QTc interval?
EMA Sebivo renewal assessment report
QTcF and concentration-response · Healthy volunteers in telbivudine QT study
QTcF changes did not exceed 5 ms, upper one-sided 95% confidence bounds did not exceed 10 ms, and QTcF did not increase with plasma telbivudine.
600 mg and 1800 mg
Cardiac safety / QTc study summary
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
85. Terodiline — Effect Predicted
The model prolongs duration with concentration, agreeing with the measured clinical QTc concentration relationship. Absence of model EADs does not validate or negate clinically observed TdP.
Clinical total plasma and stereoselective exposure are different from fixed model free concentrations. The report's QTc units and correction were retained as clinical QTc; no conversion into cellular APD90 measurements.
Concentration dependent cardiotoxicity of terodiline in patients treated for urinary incontinence
QTc and QT dispersion · Twelve treated patients in sinus rhythm with on/off-drug ECG comparisons
Mean QTc increased from 443 to 491 ms during treatment; mean drug-induced increase was 48 ms with a 95% interval of 23–74. QTc change correlated with total plasma concentration.
Stable oral treatment; measured total plasma parent and enantiomers
Primary article abstract Results; DOI 10.1136/hrt.74.1.53
Stereoselective cardiotoxic effects of terodiline
QTc · Nine healthy volunteers; randomized double-blind crossover
Racemic and R(+)-terodiline increased QTc; the S(−) enantiomer did not. Peak mean change with racemate was 23 ms at eight hours.
Racemic terodiline hydrochloride 200 mg and individual enantiomers
Primary article abstract Results
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
86. Thioridazine — Effect Predicted
Modeled APD prolongation and suspected EADs are consistent in direction with published experimental repolarization prolongation and EAD observations.
Drolet1999 used guinea-pig hearts paced at 150–250 ms, whereas CARDIX uses adult human ventricular ORd at 1000 ms. Liu2020 used chronic 24 h hiPSC exposure and mouse hearts; chronic ROS/protein-loss/calcium-channel activation mechanisms and clinical TdP are not reproduced by static inhibition. Most modeled doses overlap the published 0.3–3 µM range; 3.92 µM exceeds its upper bound. The modeled 2.94 µM point is near, but not identical to, the separate 3 µM hiPSC exposure.
Drolet et al. 1999
Monophasic action-potential duration at 90% repolarization (MAPD90) · 32 isolated buffer-perfused guinea-pig hearts
Concentration-dependent prolongation
Pacing cycle lengths 150–250 ms; greater prolongation at slower pacing.
Abstract
Liu et al. 2020
APD50/APD90; EADs in mouse hearts · Mouse hearts and human induced pluripotent stem cell cardiomyocytes
APD prolongation; increased EAD occurrence in mouse hearts
hiPSC-CM current clamp after 24 h exposure at 3 µM, n=5–8; mouse-heart optical mapping, n=5.
Abstract and Figure 8 caption
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
87. Toremifene — Effect Predicted
The positive model duration trend agrees with controlled clinical dose-related QTc prolongation; the much larger observed QTc magnitude is not claimed as quantitatively reproduced.
Parent and N-demethyl metabolite contribute clinically; model uses a fixed parent-drug profile. No unsupported oral dose-to-free-µM mapping was used for numerical experimental plotting.
FARESTON prescribing information: QTc prolongation in healthy male volunteers
Placebo- and baseline-adjusted QTc · Healthy men aged 18–45; randomized double-blind parallel study
Mean QTc increases were 7, 26 and 65 ms at 20, 80 and 300 mg; effects were dose- and concentration-related.
20, 80 and 300 mg; day 5
Section 12.2, Effects on Cardiac Electrophysiology, Table 1
Fareston product information: clinical QT study
QTc · Five-arm clinical QT study with 250 male participants
The measured QTc response showed a dose-dependent prolonging effect, with a clear positive effect in the 80 mg group.
20, 80 and 300 mg toremifene, moxifloxacin and placebo
Section 4.4, QT clinical study
No verified numeric concentration–response coordinates are available in this report; the linked source supports the qualitative comparison.
Model outputs by study and concentration
88. Voriconazole — Effect Predicted
The positive duration direction and absence of EADs agree with measured voriconazole-treated rabbit hearts. The experimental 30 µM concentration lies close to the model's highest dose; magnitude equality is not claimed.
Observed numeric values are QT changes, not APD90 values; they remain labeled as such. No numerical value was invented for 10 µM, for which the accessible abstract did not give a QT-change value. Failure to elicit EADs in one model or protocol is not a clinical safety guarantee.
Divergent electrophysiologic profile of fluconazole and voriconazole in an experimental whole-heart model of proarrhythmia
QT and EAD / polymorphic ventricular tachycardia occurrence · Isolated female rabbit whole hearts, six per drug
Voriconazole prolonged QT by 10 ms at 30 µM and 20 ms at 50 µM. No EADs or polymorphic ventricular tachycardia were induced in voriconazole-treated hearts during the sensitizing challenge.
Voriconazole 10, 30 and 50 µM; additional hypokalemia/bradycardic challenge
Primary article abstract Results; DOI 10.1016/j.ejphar.2016.02.051
Voriconazole prescribing information: cardiac electrophysiology
Placebo-adjusted QTc · Healthy men and women; randomized crossover
Mean maximal QTc increases were below 10 ms at all tested doses, with no QTc above 500 ms or increase at least 60 ms.
Single oral 800, 1200 and 1600 mg
Cardiac Electrophysiology section
Observed values by source and concentration
Model outputs by study and concentration
