AyFlux

Residual kidney function over time

This graph displays your entered/interpolated CrCl alongside model-derived acyclovir renal clearance. They are different quantities, both in mL/min. Acyclovir and CMMG renal clearance are exactly zero when the entered CrCl is zero. The linear CrCl trajectory is a user-selected exploratory assumption, not a prediction of kidney recovery or decline. [1]

Dialysis-only acyclovir clearance

Net dialysis transfer is zero outside active HD, PD dwells, and CRRT. The peritoneal transfer component may be negative during back-diffusion. Colored comparison lines can remain above zero if other scenarios with active dialysis are overlaid. [1][2]

Total acyclovir clearance (including the patient's kidneys)

Displayed instantaneous net elimination = residual kidney + nonrenal/metabolic + net dialysis transfer. During PD, net drug back-diffusion can reduce apparent total elimination. It normally remains above zero when dialysis is off because the body still eliminates some acyclovir.

Dialysis drug removal / peritoneal transfer

Membrane transfer rate is zero outside dialysis windows; PD mass is counted as removed only after a drain. Back-diffusion can make net transfer negative during a dwell. Urea Kt/V is not substituted for acyclovir clearance [1][2].

Peritoneal dialysis membrane transport

Model-predicted net ultrafiltration and dialysate acyclovir, reflecting glucose gradient decay, membrane transport and completed drains. This is a simplified mass-balance teaching model, not a validated three-pore model. [12]

Applies dialysis windows and PD exchanges together; custom medication doses are applied if edited. Otherwise the repeating dose regimen remains in effect.

Enter exact treatment times from the beginning of the simulation. For example, HD from 18 to 24 hours, or PD from 0–24 followed by HD from 30–34. Sequential HD, PD, and CRRT treatments are allowed; simultaneous modalities are not modeled.

Load an example:PD → HDHD at 18–24 h

Timeline preview

Individual PD exchange prescription

This table supports separate prescriptions for each custom PD window. Select a specific PD session or “Each PD window” for a repeated template. Exchange start times are offsets in hours from the start of that PD session, not absolute simulation hours. Apply the custom timeline to commit medication doses, dialysis windows and PD exchanges together. Measured net UF is optional; if provided, the model uses it as the net fluid-volume change over the dwell. It does not establish measured acyclovir transport or membrane sieving.

For cycler PD, start times are relative to cycler treatment start and can include turnover gaps. The table repeats with each PD session. Dialysate solute transfer remains a research approximation; observed net UF constrains water balance, not acyclovir clearance. [12]

Bayesian model updating & 90% interval coverage

How it works: Enter actual assay measurements from the same dosing and dialysis scenario. Rows marked Fit update the relative probabilities of the model's sampled parameter sets using a Gaussian assay-error assumption. Hold-out rows are never used for fitting and test whether the 90% posterior predictive interval contains each observed value. The posterior is an importance-sampling approximation, not a clinically validated population model. Do not enter identifiable patient information.

Assessment requires independent held-out measurements. Fitted data cannot validate coverage. With only a few hold-outs the coverage percentage is statistically unstable. An assay-error CV is assumed; the actual laboratory error structure may differ.

Held-out concentration diagnostics

Predictions are weighted median concentrations at each sampling time. Positive errors indicate overprediction.

Next measured concentration

Choose a cutoff after applying measurements. Both forecasts use the specified dosing schedule, dialysis schedule and renal profile. The updated forecast uses Fit rows through the cutoff. The first subsequent sampling time provides the evaluation target.

Drug reference: antiviral activity and interpretation

Herpes simplex (HSV-1 and HSV-2) and varicella-zoster virus (VZV): Acyclovir is phosphorylated first by viral thymidine kinase and then by cellular enzymes to acyclovir triphosphate, which inhibits viral DNA polymerase and terminates viral DNA synthesis [11]. This simulator predicts acyclovir and CMMG concentrations; it does not simulate viral replication, intracellular acyclovir triphosphate, clinical response, or clearance of infection.

Laboratory susceptibility: Reported cell-culture IC₅₀ ranges are HSV-1, 0.02–13.5 mg/L; HSV-2, 0.01–9.9 mg/L; and VZV, 0.12–10.8 mg/L [11]. The selected virus determines the optional laboratory susceptibility range. It does not change the modeled distribution or elimination of acyclovir. These ranges do not establish a therapeutic plasma or CSF target.

CNS infection: IV acyclovir is used in the treatment of HSV encephalitis and severe VZV disease according to indication-specific guidance. Dose, duration and renal adjustment depend on the clinical context; oral valacyclovir predictions should not be interpreted as evidence of equivalence to IV therapy in encephalitis [9] [14].

Half-life and concentration references

Published measurements grouped by kidney function and dialysis modality. Most come from small studies; use them to sanity-check model output, not as targets.

Therapeutic concentrations: No validated plasma or CSF target concentration has been established for HSV or VZV CNS infection. Neurotoxicity: CMMG 2.58 mg/L (10.8 µmol/L) is an observational risk-associated concentration, not an absolute diagnostic threshold [3].

Methods

Full differential equations, units, and parameter provenance: Read the implemented ODE system and inspect the evidence and assumption table. The formulas include oral absorption, CMMG, PD mass transfer, time-varying CrCl, and separate CSF lag updates.

Acyclovir (whether administered IV or generated from oral valacyclovir) is represented with a two-compartment plasma/peripheral distribution model, separate exploratory CSF effect outputs, and a formed CMMG compartment. Residual renal clearance and time-dependent dialysis clearance act on circulating drug. The model uses representative literature dialysis values as anchors, but most remaining parameters have not been estimated from patient-level data in an externally validated population PK model [1][2].

Zero residual kidney function: both the acyclovir and CMMG renal elimination terms are multiplied by CrCl and are exactly zero at CrCl 0 mL/min, including every Monte Carlo draw. Nonrenal metabolism and active dialysis clearance remain separate, and may stay positive. The plotted renal-only clearance is separate from total drug clearance. This mathematical property is not a verified estimate of all clearance pathways in an individual patient.

Medication schedules: The repeating regimen generates no more than the selected total number of administrations, beginning at hour 0 and spaced by the chosen interval. Administrations scheduled beyond the displayed horizon are not simulated. The custom medication timeline supports a sequence of oral valacyclovir and IV acyclovir administrations at specific hours; an IV dose is a timed infusion, while oral valacyclovir enters a first-order absorption depot. The same plasma, peripheral, CSF and CMMG amounts persist across administration-route changes and dialysis modalities. Adding or changing medication routes does not reset drug concentrations. Oral and IV dose milligram strengths must not be treated as interchangeable. This is an exploratory mathematical implementation and should not be used for clinical dosing [8] [9].

Oral valacyclovir: tablets are represented with a simplified first-order oral depot followed by rapid conversion to acyclovir. The active plasma drug is acyclovir, and CMMG is formed from it. The assumed fraction of acyclovir entering circulation after valacyclovir is 0.545, anchored to the US label's healthy-volunteer 1-g study. Dose strengths are valacyclovir free-base milligrams; conversion uses the free-base molecular-weight ratio 225.21/324.34. Absorption-rate and relative bioavailability distributions are illustrative rather than fitted; the label notes less-than-dose-proportional exposure across 250–1000 mg [8]. For exploratory input, oral valacyclovir accepts up to 6,000 mg per administration; this is not a dose recommendation. For >1,000 mg, the simplified constant-bioavailability model extrapolates beyond its anchor and may overestimate exposure due to nonlinear absorption. These oral predictions must not be used to replace IV treatment for suspected HSV or VZV encephalitis [9][24].

HD: an illustrative dialytic clearance anchored to the mean 81.8 mL/min in a small 1982 study [1]. PD: the historical CAPD study reported 3.4 mL/min measured total peritoneal clearance [2]. The simulator now uses a separate assumed membrane transfer coefficient, integrates diffusive and small-pore convective acyclovir transfer into a dialysate drug-mass compartment, tracks fluid volume and drains at exchange boundaries. Glucose concentration changes a simplified time-decaying osmotic ultrafiltration term; aquaporin water-only transport does not directly carry acyclovir. The illustrative transport, glucose-decay, reabsorption and reflection parameters are not fitted to drug concentrations [12]. A PD-to-HD toxicity case supplies clinical context but cannot validate those transport terms [23]. Dialysate-associated drug is counted as removed when drained, not automatically at the time of membrane transfer. CMMG PD elimination remains a separate exploratory approximation.

Cockcroft–Gault CrCl: Estimated CrCl (mL/min) = (140 − age) × actual weight (kg) × sex coefficient / [72 × serum creatinine (mg/dL)]; sex coefficient 1.0 for male or 0.85 for female. Creatinine-based estimates are unreliable with unstable kidney function, dialysis, or atypical muscle mass; measured residual clearance may be preferred [13].

Time limits: each user-entered interval is at most 24 hours; ongoing PD or CRRT is entered as successive daily intervals. No treatment overlap is allowed. Residual CrCl is expressed in absolute mL/min. When Cockcroft–Gault is selected, the sex coefficient and actual body weight are used for estimation. The concentration bands are prior parameter ranges, or posterior-weighted parameter ranges after optional measurement fitting, and are not established clinical prediction errors.

CNS: plasma–CSF exchange and CMMG entry are a lagged partition approximation, not a physiologically based permeability model. CNS drug concentrations, viral susceptibility, and neurotoxicity markers should not be used to identify a suitable patient dose [3][4][5][19][24]. Measured studies in different populations found comparatively flat steady-state CSF acyclovir levels and variable CSF levels during CNS infection; these are important research constraints, not a validated patient-specific blood–brain barrier model [17][18].

Model half-life: A single half-life is not displayed in the modality comparison because dynamic dialysis schedules do not have a single meaningful elimination half-life. The comparison instead reports the predicted plasma concentration at the end of the simulation. The model may internally calculate an off-dialysis terminal eigenvalue for research diagnostics, but this is not a validated clinical half-life. Published observed half-lives are listed in the drug reference above.

Bayesian 90% bands: The implementation uses weighted importance sampling of 120 default parameter draws. Entered fit measurements update each draw by a Gaussian likelihood (assay CV adjustable in Advanced Settings), and a separate held-out measurement is compared with the 90% posterior predictive range including assumed measurement error. Posterior parameter intervals and predictive intervals are different. Uncalibrated assumed priors and small effective sample sizes limit validity; there is no claim of 90% real-world coverage without external measured data.

Peritoneal dialysis (PD) exchange table: The optional prescription table lets you describe each exchange separately. Dwell is how long the fluid remains in the abdomen; fill is the volume of dialysis fluid added; dextrose describes the fluid concentration; and net ultrafiltration is the estimated amount of fluid removed (optional). These entries affect a simplified fluid and drug-transfer model, not a validated prediction of clearance for an individual patient.

Research analyses: The optional model-fitting comparison uses published average peak and trough values for an assumed 70-kg adult with CrCl 95 mL/min. It is not a patient-specific fit or independent validation. Sensitivity analysis changes one parameter at a time by ±20% to show its effect on simulated results. Configuration files can be exported and imported locally in the browser.

References

Sources are organized by their historical reference numbers. They include measured pharmacokinetic cohorts, case reports, labeling, practice guidance, and clinical commentary. A citation provides scientific context but does not mean that AyFlux parameters have been fitted or clinically validated against that source.

Open the complete bibliography on its own page · Cite the software

1. Laskin OL, et al. Acyclovir kinetics in end-stage renal disease. Clin Pharmacol Ther. 1982;31(5):594–601. doi:10.1038/clpt.1982.83. Two-compartment kinetics and HD clearance.
2. Burgess ED, Gill MJ. Intraperitoneal administration of acyclovir in patients receiving continuous ambulatory peritoneal dialysis. J Clin Pharmacol. 1990;30(11):997–1000. doi:10.1002/j.1552-4604.1990.tb03585.x. Measured peritoneal clearance.
3. Helldén A, et al. High serum concentrations of the acyclovir main metabolite 9-carboxymethoxymethylguanine in renal failure patients with acyclovir-related neuropsychiatric side effects: an observational study. Nephrol Dial Transplant. 2003;18(6):1135–1141. doi:10.1093/ndt/gfg119. Observed CMMG association, not a validated diagnostic threshold.
4. Lindström J, et al. An unexpectedly high occurrence of aciclovir-induced neuropsychiatric symptoms in patients treated for herpesvirus CNS infection: a prospective observational study. J Antimicrob Chemother. 2019;74(12):3565–3572. doi:10.1093/jac/dkz357. CSF and serum observations and barrier effects.
5. US acyclovir oral labeling. VZV in-vitro IC₅₀ range and important exposure limitations.
6. Boelaert J, et al. Multiple dose pharmacokinetics of intravenous acyclovir in patients on continuous ambulatory peritoneal dialysis. J Antimicrob Chemother. 1987;20(1):69–76. doi:10.1093/jac/20.1.69. Fitted three-compartment plasma, peripheral and peritoneal-dialysate kinetics; numerical parameters require careful model-specific matching.
7. ISPD practice guidelines and resources. PD prescription dimensions including modality, dwell/exchange parameters and dialysate solutions.
8. VALTREX prescribing information, clinical pharmacology. Valacyclovir is rapidly converted to acyclovir; reported mean absolute acyclovir bioavailability 54.5% ± 9.1% after 1 g oral dose; tablets labeled in valacyclovir free-base equivalents; oral exposure is not strictly dose proportional.
9. NIH VZV guidance for adults and adolescents with HIV. Illustrates separate indications for oral antivirals in localized zoster versus IV acyclovir in severe/complicated infections.
10. Acyclovir injection prescribing information: normal half-life, steady-state peak and trough, renal impairment.
11. Oral acyclovir prescribing information: HSV-1, HSV-2, and VZV in-vitro IC₅₀ ranges, antiviral mechanism, and absence of a validated susceptibility-to-outcome relationship.
12. Morelle J, et al. ISPD recommendations for the evaluation of peritoneal membrane dysfunction in adults: Classification, measurement, interpretation and rationale for intervention. Perit Dial Int. 2021;41(4):352–372. doi:10.1177/0896860820982218. osmosis, small-solute transport, glucose gradient and ultrafiltration. Values used in this website's PD transport submodel are exploratory, not fit to an acyclovir APD dataset.
13. National Kidney Foundation: Cockcroft–Gault equation and limitations. Historical CrCl estimator; accuracy varies with body composition, age and creatinine kinetics.
14. NIH HSV adult/adolescent opportunistic infection guidelines: HSV CNS infection management and distinction between IV and oral antiviral regimens.
15. Smith JP, et al. Pharmacokinetics of acyclovir and its metabolites in cerebrospinal fluid and systemic circulation after administration of high-dose valacyclovir in subjects with normal and impaired renal function. Antimicrob Agents Chemother. 2010;54(3):1146–1151. doi:10.1128/AAC.00729-09. Measured CSF/plasma AUC ratios; acyclovir ~0.25, CMMG ~0.025 in a small study.
16. Shah GM, Winer RL, Krasny HC. Acyclovir pharmacokinetics in a patient on continuous ambulatory peritoneal dialysis. Am J Kidney Dis. 1986;7(6):507–510. doi:10.1016/S0272-6386(86)80193-7. Reported peritoneal clearance 4.4 mL/min; single-patient evidence.
17. Lycke J, Malmström C, Ståhle L. Acyclovir levels in serum and cerebrospinal fluid after oral administration of valacyclovir. Antimicrob Agents Chemother. 2003;47(8):2438–2441. doi:10.1128/AAC.47.8.2438-2441.2003. Measured steady-state oral valacyclovir serum and CSF acyclovir in 10 participants with MS, without documented blood–CSF barrier disruption; provides concentration-time and CSF-to-serum AUC comparisons, not direct validation of AyFlux.
18. Pouplin T, et al. Valacyclovir for herpes simplex encephalitis. Antimicrob Agents Chemother. 2011;55(7):3624–3626. doi:10.1128/AAC.01023-10. Small observational oral-valacyclovir CNS-infection cohort with serial plasma and CSF acyclovir measurements; treatment efficacy and equivalence to IV acyclovir were not established.
19. Helldén A, et al. The aciclovir metabolite CMMG is detectable in the CSF of subjects with neuropsychiatric symptoms during aciclovir and valaciclovir treatment. J Antimicrob Chemother. 2006;57(5):945–949. doi:10.1093/jac/dkl067. Retrospective comparison of symptomatic and asymptomatic subjects; CMMG association is not proof of causation or a validated toxicity threshold.
20. Lycke J, Andersen O, Svennerholm B, Appelgren L, Dahlöf C. Acyclovir concentrations in serum and cerebrospinal fluid at steady state. J Antimicrob Chemother. 1989;24(6):947–954. doi:10.1093/jac/24.6.947. Measurements after oral acyclovir in MS; useful for CSF kinetics context but not a direct AyFlux route benchmark (oral acyclovir is not implemented).
21. Almond MK, Fan S, Dhillon S, Pollock AM, Raftery MJ. Avoiding acyclovir neurotoxicity in patients with chronic renal failure undergoing haemodialysis. Nephron. 1995;69(4):428–432. doi:10.1159/000188514. Oral acyclovir concentrations and HD removal in seven oliguric patients; the studied oral-aciclovir route is not available in AyFlux.
22. Vasquez Vasquez JA, et al. Neurotoxicity from inordinate valacyclovir dosage in an elderly woman with stage 5 chronic kidney disease. Br J Clin Pharmacol. 2025;91(8):2469–2473. doi:10.1002/bcp.70138. Published single case with blood, dialysate and urine measurements; body-burden estimates remain model-dependent. The 2024 preprint was not peer-reviewed; it is retained as publication history; the current HD reconstruction cites the final journal article. Publication does not create a second independent case or validate AyFlux.
23. Sadjadi S-A, Regmi S, Chau T. Acyclovir neurotoxicity in a peritoneal dialysis patient: report of a case and review of the pharmacokinetics of acyclovir. Am J Case Rep. 2018;19:1459–1462. doi:10.12659/AJCR.911520. PD-to-HD clinical case, not prospective dose-response validation or a neurotoxicity cutoff.
24. Bodilsen J, Nielsen H, Whitley RJ. Valaciclovir therapy for herpes encephalitis: caution advised. J Antimicrob Chemother. 2019;74(6):1467–1468. doi:10.1093/jac/dky568. Clinical commentary cautioning against inferring HSV encephalitis treatment efficacy from CSF concentrations or in-vitro inhibitory concentrations; not an independent PK validation cohort.
25. Abdalla S et al. Population Pharmacokinetics of Intravenous and Oral Acyclovir and Oral Valacyclovir in Pediatric Population To Optimize Dosing Regimens. Antimicrob Agents Chemother. 2020;64:e01426-20. doi:10.1128/AAC.01426-20. Source for the separate pediatric IV research module; not validation of the adult engine or pediatric dialysis.
26. Zeng L, et al. Population pharmacokinetics of acyclovir in children and young people with malignancy after administration of intravenous acyclovir or oral valacyclovir. Antimicrob Agents Chemother. 2009;53(7):2918–2927. doi:10.1128/AAC.01138-08. 43 patients; final Table 3 supports a separate pediatric IV comparison. Counahan renal estimates and L/h/m² units differ from Abdalla’s historical Schwartz input.
27. Soul-Lawton J, et al. Absolute bioavailability and metabolic disposition of valaciclovir, the L-valyl ester of acyclovir, following oral administration to humans. Antimicrob Agents Chemother. 1995;39(12):2759–2764. doi:10.1128/aac.39.12.2759. Mean absolute bioavailability 54.2% in 12 healthy volunteers. AyFlux’s 54.5% default remains attributed to the product label, not this result.
28. Weller S, et al. Pharmacokinetics of the acyclovir pro-drug valaciclovir after escalating single- and multiple-dose administration to normal volunteers. Clin Pharmacol Ther. 1993;54(6):595–605. doi:10.1038/clpt.1993.196. Human dose-escalation evidence for less-than-dose-proportional oral exposure; does not determine a universal saturation constant.
29. Kably B, et al. Population pharmacokinetics of aciclovir and its major metabolite 9-carboxymethoxymethylguanine and safety profile of valaciclovir in early liver transplant recipients. J Antimicrob Chemother. 2025;80(5):1302–1308. doi:10.1093/jac/dkaf070. 50 transplant recipients; only one participant had eGFR below 30. Joint model and simulations support renal-dependence research, not dialysis validation or a toxicity threshold.
30. Stathoulopoulou F, et al. Evaluation of Valaciclovir Dosage Reduction in Continuous Ambulatory Peritoneal Dialysis Patients. Nephron. 2002;91(1):164–166. doi:10.1159/000057621. 12 participants after one 500 mg dose; only eight had three timed plasma samples. Apparent clearance is approximate; full AUC, bioavailability and dialysis clearance were not established.
31. Stathoulopoulou F, et al. Clinical Pharmacokinetics of Oral Acyclovir in Patients on Continuous Ambulatory Peritoneal Dialysis. Nephron. 1996;74(2):337–341. doi:10.1159/000189332. Oral acyclovir in CAPD; do not transfer formulation-dependent apparent parameters directly to valacyclovir or IV dosing.
32. Kacirova I, et al. Therapeutic monitoring of serum concentrations of acyclovir and its metabolite 9-(carboxymethoxymethyl) guanine in routine clinical practice. Biomed Pharmacother. 2022;156:113852. doi:10.1016/j.biopha.2022.113852. 27 patients receiving IV acyclovir. Renal strata differed in dose and demographics; pooled troughs are observational comparators, not a controlled dose-response experiment.
33. Funaki T, et al. Therapeutic Drug Monitoring in Neonatal HSV Infection on Continuous Renal Replacement Therapy. Pediatrics. 2015;136(1):e270–e274. doi:10.1542/peds.2014-3380. One neonate with changing CRRT and plasma exchange. Circuit and treatment timing prevent use as a general neonatal dialysis-clearance constant.
34. D’Agate S, et al. Population pharmacokinetics and dose rationale for aciclovir in term and pre-term neonates with herpes. Pharmacol Res Perspect. 2024;12(3):e1193. doi:10.1002/prp2.1193. Final analysis: 83 samples from 28 infants. Reanalysis of earlier neonatal trial data; not an independent validation cohort for the original model.
35. Sun M, et al. Revisiting Acyclovir Dosing for Adult Viral Encephalitis Using a Full Bayesian LeiCNS PBPK Modeling Approach. Clin Pharmacokinet. 2025;64(9):1413–1423. doi:10.1007/s40262-025-01545-w. CNS PBPK analysis distinguishes CSF from brain extracellular fluid. Reuses published observations; model code is available on reasonable request, not supplied in this PDF.
36. Laskin OL, et al. Effect of Renal Failure on the Pharmacokinetics of Acyclovir. Am J Med. 1982;73(1A):197–201. doi:10.1016/0002-9343(82)90090-0. Six anuric participants; likely overlapping with the other 1982 Laskin/Krasny reports. Treat as potentially shared evidence, not three independent cohorts.
37. Krasny HC, et al. Influence of Hemodialysis on Acyclovir Pharmacokinetics in Patients with Chronic Renal Failure. Am J Med. 1982;73(1A):202–204. doi:10.1016/0002-9343(82)90091-2. Historical dialysis experiment. The reported concentration reduction is not the fraction of the initial dose removed; match timing and the amount denominator before comparison.
38. Boulieu R, Bastien O, Gaillard S, Flamens C. Pharmacokinetics of acyclovir in patients undergoing continuous venovenous hemodialysis. Ther Drug Monit. 1997;19(6):701–704. doi:10.1097/00007691-199712000-00016. Three adult ICU patients on CVVHD; half-life, ultrafiltrate clearance and sieving coefficient. Abstract-level values; full text not reviewed.
39. Bleyzac N, Barou P, Massenavette B, et al. Assessment of acyclovir intraindividual pharmacokinetic variability during continuous hemofiltration, continuous hemodiafiltration, and continuous hemodialysis. Ther Drug Monit. 1999;21(5):520–525. doi:10.1097/00007691-199910000-00005. One pediatric patient; Bayesian estimates from routine monitoring. Abstract-level, qualitative.
40. Collignon C, de Marcellus C, Oualha M, et al. Pharmacokinetic profile of acyclovir in a child receiving continuous kidney replacement therapy for acute liver failure. Pediatr Nephrol. 2023;38(10):3493–3497. doi:10.1007/s00467-023-05881-6. Single 21-month-old on CVVHDF; open access.
41. Laskin OL, de Miranda P, King DH, et al. Effects of probenecid on the pharmacokinetics and elimination of acyclovir in humans. Antimicrob Agents Chemother. 1982;21(5):804–807. doi:10.1128/AAC.21.5.804. Three adults with normal kidney function; Table 1 serial plasma means (5 mg/kg IV) are the study overlay and scorecard data. Not used to fit any parameter.
42. Kacirova I, Urinovska R, Sagan J. The importance of therapeutic drug monitoring in dosage optimization of acyclovir. Klinicka farmakologie a farmacie. 2022. Figure 1 is the source of the paired single-patient acyclovir and CMMG observations; distinct from the 27-patient cohort in [32].
43. Khajehdehi P, Jamal JA, Bastani B. Removal of acyclovir during continuous veno-venous hemodialysis and hemodiafiltration with high-efficiency membranes. Clin Nephrol. 2000;54(4):351–355. PMID 11076113 (no DOI). One anuric adult (92–94 kg) on CVVHD and CVVHDF; measured prefilter, postfilter and effluent concentrations with derived clearances. Table 1 labels flows in mL/min, but the text and arithmetic show mL/h.
44. Moysés Neto M, Nakagawa B, Chiozi SZ, da Motta FU, da Silva Junior JC, Gonçalves WE. Neurotoxicity from acyclovir and valacyclovir in patients undergoing peritoneal dialysis: still rare, but still present. Perit Dial Int. 2026. doi:10.1177/08968608261415988. Three automated (cycler) PD patients with doses and cycler prescriptions; no drug concentrations were measured.

This simulator is for model comparison and teaching. An optional finite-ensemble Bayesian reweighting tool accepts investigator-entered concentrations, but no real patient data are supplied, and no independent external clinical validation has been performed. Oral valacyclovir results have not been validated for CNS treatment decisions.

Evidence scorecard

How closely the default model reproduces published concentrations it was not fitted to. For each comparison scenario, AyFlux runs the published regimen, including any dialysis, and compares its nominal curve with each reported value. Predicted/observed is the geometric mean ratio across points (1.00 is perfect; 0.77 means 23% low on average). Inside model interval counts reference values within assumed 90% parameter-variability bands from 400 draws. These are not calibrated predictive intervals for independent patient measurements. Group means are not individual patients; single-patient reports depend on reconstructed timing; Boelaert values are the authors' fitted peaks and troughs. Weight, timing and kidney function are nominal where studies did not report them. Calibration multipliers from Research tools carry over.

Predicted versus observed

Errors by sample type and time since dose

Mass recovery comparison

Stathoulopoulou 2002 reports urinary and CAPD recovery over 24 hours. These amount endpoints are separate from the concentration scorecard and its pooled metrics. Inspect published recovery, denominator ambiguity and model sensitivity.

Experimental

Early research modules with their own published models and evidence limits.

AyFlux Neonatal

Explore predicted IV acyclovir clearance from 25 to 41 weeks postmenstrual age. Reproduces the published Sampson infant equation; no dose recommendations.

Guided scenarios

Each button loads a teaching scenario and opens the Concentration tab with a short note on what to look at. Your current settings are saved to version history first, so Undo last import brings them back. The regimens are deliberately unadjusted illustrations, not recommendations.

Save / load a simulation

Export or import a local JSON configuration, including patient assumptions, dosing, dialysis schedules, exchange prescriptions, and model parameters. No information is sent to a server. Do not store identifiable patient data.

Compare an external PK-Sim / MoBi CSV

Have an XLSX workbook? Open it in Excel or LibreOffice, select the worksheet containing one concentration series, and save that sheet as CSV. Choose that CSV below, confirm the time and concentration columns, units and target analyte, then click Compare curves. XLSX import is not supported here.

Import simulation results, not native model files. Everything stays in your browser. Choose the exported columns and units explicitly; no clinical comparison or validation is implied. Data exchange guide.

Select a CSV to begin.

Drug flow and 3D visual explorer

Follow modeled drug amounts through a Sankey diagram, explore two-input heatmaps, and open an optional 3D response surface. Select a scenario to view its concentration curves.

Copies the current successfully simulated settings. Opens in a separate page.

Two-parameter sensitivity heatmap

Exploratory deterministic sweep over residual creatinine clearance and intermittent hemodialysis clearance. Each cell reruns the same simulated dose schedule, with a single parameter set. Other settings, including the full simulation duration, are held fixed. Not a dose selection chart.

Hemodialysis timing explorer

Moves one hemodialysis session of the current length across the simulation and reruns the current dose schedule each time, with a single nominal parameter set. It shows how the timing of a session relative to doses changes the drug removed and the exposure. Session start times cover the first 96 hours; the simulation keeps its full duration, kidney-function trajectory and dose schedule. Other settings are held fixed. Not a dialysis scheduling tool.

Numerical integrity checks

Checks the current prediction arrays, zero-renal-clearance edge case when applicable, and a deterministic RK4 time-step comparison. Mass balance is an approximate drug-amount accounting diagnostic. None of these checks establishes clinical validity.

Research report (PDF)

Create a locally generated, print-ready report with model assumptions, summary metrics, scientific limitations, numerical diagnostics, and available chart images. Choose Save as PDF in your browser's print dialog. No model data are uploaded.

Published concentration benchmark

The US IV acyclovir label reports mean steady-state plasma peaks and troughs at 5 mg/kg (n=8) and 10 mg/kg (n=7). [10] AyFlux simulates ten 1-hour infusions of 10 mg/kg given every 8 hours. The 5 mg/kg row uses half the predicted peak and trough from that simulation.

Historical anuric benchmark (not fitted)

Laskin et al. (1982): six anuric HD patients received a single 2.5 mg/kg IV infusion over 1 hour, with measurements before dialysis. Literature mean ± SD values are converted from µmol/L to mg/L using acyclovir molecular weight 225.21 g/mol. Model assumptions: 70 kg, CrCl 0 mL/min, no HD during the first 24 hours. Timing and weight are nominal approximations, not patient-matched data. [1]

Fitting minimizes squared log errors for group mean peaks/troughs by adjusting central distribution volume and renal/nonrenal plasma clearance. The model cannot perfectly reproduce all published group means by a two-parameter linear fit. This is calibration against the same summary statistics shown above—not external validation, and not supported for individualized or dialysis prescribing.

One-at-a-time sensitivity analysis

Quantifies the influence of selected assumptions on drug exposure. The table shows percentage change in an output when the input is decreased by 20% / increased by 20%. Results use nominal parameter values, not fitted or posterior uncertainty bands.

Replay published study cases

Import the downloaded JSON through Load configuration to explore the scenario in the main simulator. Smith cases compare the default CMMG CSF coefficient of 0.25 with an exploratory value of 0.025. These comparisons use the same study that informed the alternative.