Tutorials › Pharmacometrics › QSP Models of Hemodynamics
Quantitative Systems Pharmacology · Cardiovascular Modeling

QSP Models of Hemodynamics

Learn how quantitative systems pharmacology models represent blood flow, pressure, vascular resistance, cardiac output, blood volume, and cardiovascular drug effects as interacting components of a mechanistic physiological system.

Intermediate QSP Modeling Cardiovascular Systems Hemodynamics
01 · The big picture

1. What Is Hemodynamics?

Hemodynamics describes the movement of blood through the cardiovascular system and the physical relationships among pressure, flow, vascular resistance, cardiac output, and blood volume.

In quantitative systems pharmacology (QSP), hemodynamics provides a mechanistic framework for connecting a drug's molecular or pharmacologic effects to changes in cardiovascular physiology. Rather than treating blood pressure as an isolated endpoint, a QSP model can represent blood pressure as an emergent consequence of interacting physiological processes.

Drug exposure / effect Hemodynamic system heart · vessels blood volume · resistance BP & Flow MAP · CO · SV physiological feedback: baroreflex and homeostasis

A cardiovascular QSP model links drug effects to interacting hemodynamic processes rather than modeling a blood-pressure endpoint in isolation.

Core idea: hemodynamic QSP models treat cardiovascular variables as components of a dynamic system. Changes in vascular tone, cardiac function, blood volume, or autonomic regulation can propagate through the system and alter pressure and flow.
02 · What QSP asks

2. What Questions Can a Hemodynamic QSP Model Answer?

A hemodynamic QSP model can address questions that require integration across multiple physiological processes.

QuestionModel componentWhat it helps describe
What determines arterial pressure?MAP, cardiac output, vascular resistanceThe interaction between pressure generation and systemic vascular load
How does a drug lower blood pressure?Pharmacodynamic mechanismChanges in cardiac function, vascular tone, fluid balance, or combinations of these
Why does heart rate change after treatment?Autonomic regulationBaroreflex-mediated and direct drug effects on heart rate
How does vascular resistance affect flow?Resistance-flow relationshipsThe dependence of pressure gradients and flow on vascular tone
Why can blood pressure recover after an acute perturbation?Feedback controlHomeostatic responses such as baroreflex regulation
What happens under repeated dosing?Dynamic drug-system modelThe interaction between drug exposure, effect, and physiological adaptation

The key distinction from a purely empirical model is that the QSP framework attempts to explain the observed response through interconnected physiological mechanisms.

03 · Core variables

3. The Core Hemodynamic Variables

Most mechanistic cardiovascular models are built around a relatively small set of fundamental variables. These variables are related mathematically, but each represents a different aspect of cardiovascular physiology.

VariableMeaningRole in a QSP model
MAPMean arterial pressureRepresents the average arterial pressure driving systemic blood flow
COCardiac outputTotal blood flow generated by the heart per unit time
HRHeart rateFrequency of cardiac contractions
SVStroke volumeBlood ejected by the ventricle per beat
SVRSystemic vascular resistanceRepresents the resistance opposing systemic blood flow
BVBlood volumeProvides a state variable connecting fluid balance with filling pressures and cardiac output

A particularly important relationship is:

$$CO=HR\times SV$$

Cardiac output can therefore change because heart rate changes, stroke volume changes, or both.

Under a simplified systemic circulation representation, mean arterial pressure is related to cardiac output and systemic vascular resistance by:

$$MAP\approx CO\times SVR$$

More precisely, the pressure driving systemic flow is related to the difference between mean arterial pressure and venous pressure:

$$MAP-P_{CV}\approx CO\times SVR$$

where \(P_{CV}\) represents an appropriate central venous or downstream pressure approximation. The simpler \(MAP\approx CO\times SVR\) relationship is often useful when downstream venous pressure is small relative to arterial pressure.

04 · Pressure and flow

4. Pressure, Flow, and Resistance

At the level of a simplified vascular bed, blood flow depends on the pressure difference across the vascular system and the resistance to flow.

$$Q=\frac{\Delta P}{R}$$

Equivalently:

$$\Delta P=Q\times R$$

This relationship is analogous to other transport systems in physiology and engineering. A pressure difference provides the driving force, while vascular resistance determines how much flow results from that pressure difference.

high pressure ΔP Resistance vascular tone Q blood flow

In a simplified vascular representation, flow is determined by the pressure gradient and resistance.

For QSP modeling, the importance of this relationship is that a drug can alter pressure indirectly. For example, a vasodilator may reduce vascular resistance, which can lower arterial pressure even when cardiac output changes in response.

05 · Vascular tone

5. Vascular Resistance and Vessel Radius

For idealized laminar flow through a cylindrical tube, Poiseuille's law illustrates the strong dependence of resistance on vessel radius:

$$R=\frac{8\eta L}{\pi r^4}$$

Here, \(\eta\) is viscosity, \(L\) is vessel length, and \(r\) is vessel radius.

The fourth-power dependence on radius means that relatively small changes in vascular radius can produce large changes in theoretical hydraulic resistance. Real blood vessels are not rigid cylindrical tubes, and the circulation is substantially more complex, but the relationship provides useful mechanistic intuition for why vascular tone can have a large influence on systemic hemodynamics.

QSP implication: a drug that modifies smooth-muscle tone does not need to act directly on blood pressure. Its effect can propagate through vessel radius, resistance, flow, and pressure.
06 · Cardiac function

6. Cardiac Output, Stroke Volume, and Heart Rate

Cardiac output is the amount of blood pumped by the heart per unit time:

$$CO=HR\times SV$$

Stroke volume can be expressed as the difference between end-diastolic and end-systolic volume:

$$SV=EDV-ESV$$

Therefore:

$$CO=HR(EDV-ESV)$$

This decomposition is important for mechanistic modeling because drugs can influence cardiac output through different pathways.

MechanismPotential model variable affectedHemodynamic consequence
ChronotropyHeart rateChanges cardiac output through HR
InotropyContractility / ESVChanges stroke volume and potentially cardiac output
PreloadEDV / fillingChanges ventricular filling and stroke volume
AfterloadVascular resistance / pressureCan alter ventricular ejection and stroke volume

A QSP model can explicitly represent these pathways when the scientific question requires them, rather than treating cardiac output as a single unexplained variable.

07 · Fluid balance

7. Blood Volume and Cardiovascular Filling

Blood volume is an important state variable in many cardiovascular systems models because changes in fluid volume can influence venous return, cardiac filling, cardiac output, and arterial pressure.

A simplified mass-balance representation can be written as:

$$\frac{dBV}{dt}=Input_{fluid}-Output_{fluid}$$

The actual physiology can be expanded to include renal sodium and water handling, vascular permeability, interstitial fluid exchange, hormonal regulation, and other processes.

In a QSP model, this creates a connection between cardiovascular hemodynamics and other physiological systems. For example, a drug that changes renal sodium handling may alter extracellular fluid volume, which can subsequently affect cardiovascular filling and blood pressure.

Systems perspective: blood pressure is not determined solely by the heart and arteries. Fluid balance, renal function, neurohormonal regulation, and vascular tone can all contribute to the observed cardiovascular state.
08 · Feedback

8. The Baroreflex: A Key Homeostatic Mechanism

The cardiovascular system contains feedback mechanisms that oppose acute changes in arterial pressure. The baroreflex is one of the most important of these mechanisms.

In a simplified representation, arterial pressure is sensed by baroreceptors, and changes in the sensed pressure influence autonomic activity. Autonomic activity can then alter heart rate, contractility, and vascular tone.

Arterial pressure Baroreceptors pressure sensing Autonomic response heart rate · contractility · vascular tone

A simplified baroreflex closes a feedback loop between arterial pressure and cardiovascular effectors.

A useful conceptual representation is:

$$MAP\rightarrow Baroreceptor\ signal\rightarrow Autonomic\ activity\rightarrow HR,\ contractility,\ SVR\rightarrow MAP$$

The feedback loop means that the final effect of a drug may differ from its immediate direct pharmacologic effect. A direct decrease in vascular resistance, for example, may trigger compensatory changes in heart rate or contractility.

09 · Pharmacology

9. How Drug Effects Enter a Hemodynamic QSP Model

The pharmacologic component of a QSP model describes how drug exposure changes one or more physiological processes.

A simple direct-effect relationship might use an \(E_{\max}\) model:

$$E(C)=E_0+\frac{E_{\max}C}{EC_{50}+C}$$

Depending on the drug and mechanism, the effect could modify vascular resistance, heart rate, contractility, renal sodium handling, venous tone, or another physiological process.

For example, if drug concentration reduces vascular resistance, a simplified relationship might be:

$$SVR(C)=SVR_0\left(1-\frac{E_{\max}C}{EC_{50}+C}\right)$$

The hemodynamic model then propagates that change through the cardiovascular system.

Important distinction: the pharmacodynamic model specifies how drug exposure changes a physiological mechanism. The hemodynamic model determines how that mechanistic perturbation propagates through the cardiovascular system.
10 · Dynamic equations

10. From Physiology to Differential Equations

QSP models represent physiological processes using differential equations, algebraic equations, or combinations of both.

Suppose a simplified model contains blood volume \(BV\), cardiac output \(CO\), and systemic vascular resistance \(SVR\). A basic algebraic pressure relationship might be:

$$MAP=CO\times SVR+P_{CV}$$

If blood volume changes dynamically, a simple mass-balance equation could be:

$$\frac{dBV}{dt}=J_{in}-J_{out}$$

Cardiac output can be represented as:

$$CO=HR\times SV$$

And stroke volume can depend on physiological state:

$$SV=f(EDV,ESV,contractility,afterload)$$

These equations form a connected system rather than independent formulas. A change in one state variable can therefore propagate through several equations and ultimately alter an observable such as blood pressure.

11 · Model structure

11. Cardiovascular Compartments in QSP Models

Unlike a simple PK compartment, a cardiovascular compartment can represent a physiologically meaningful blood or tissue space. The exact structure depends on the purpose of the model.

ComponentPotential representationTypical role
HeartCardiac pumpGenerates pressure and flow
Arterial compartmentSystemic arterial blood volumeRepresents arterial pressure and compliance
Venous compartmentSystemic venous blood volumeProvides a reservoir and affects venous return
Peripheral vascular bedsResistance elementsRepresent regional vascular tone and flow
Pulmonary circulationPulmonary vascular compartmentConnects right and left cardiac output
Renal systemFluid and electrolyte regulationLinks kidney function to volume and pressure

More detailed models may subdivide the systemic circulation into multiple vascular beds, such as renal, cerebral, splanchnic, skeletal muscle, and cutaneous circulations. This can be useful when drug effects differ substantially among tissues.

12 · Vascular mechanics

12. Compliance and the Relationship Between Volume and Pressure

Blood vessels are compliant rather than perfectly rigid. Compliance describes how much volume changes for a given change in pressure:

$$C=\frac{dV}{dP}$$

For a simplified linear representation:

$$\Delta V=C\Delta P$$

Arterial compliance is particularly important because arterial pressure depends not only on flow and resistance but also on how the vascular system stores and releases blood during the cardiac cycle.

A QSP model can therefore distinguish between:

  • Resistance: opposition to blood flow.
  • Compliance: the volume response to changes in pressure.
  • Elastance: the inverse of compliance in appropriate formulations.
  • Volume: the amount of blood contained within a vascular compartment.

Including these properties allows the model to represent dynamic pressure responses rather than only steady-state relationships.

13 · Worked example

13. Worked Example: How a Change in Vascular Resistance Changes Pressure

Consider a simplified cardiovascular system with:

  • Cardiac output: 5 L/min
  • Systemic vascular resistance: 18 mmHg·min/L
  • Central venous pressure: 3 mmHg

Step 1: Calculate mean arterial pressure

Using the simplified pressure-flow relationship:

$$MAP=CO\times SVR+P_{CV}$$
$$MAP=(5)(18)+3=93\text{ mmHg}$$

Step 2: Introduce a vasodilator effect

Suppose the drug reduces systemic vascular resistance by 25%, while cardiac output initially remains at 5 L/min.

$$SVR_{new}=18(1-0.25)=13.5\text{ mmHg·min/L}$$

Step 3: Recalculate pressure

$$MAP_{new}=(5)(13.5)+3=70.5\text{ mmHg}$$

Step 4: Interpret the result

Under these deliberately simplified assumptions, a 25% reduction in systemic vascular resistance lowers mean arterial pressure from approximately 93 to 70.5 mmHg.

A real QSP model would not necessarily stop here. The fall in arterial pressure could activate baroreflex responses, alter heart rate and contractility, change venous tone, and potentially modify cardiac output. The final predicted pressure would therefore emerge from the complete interconnected model.

Why QSP matters: the simple calculation describes an immediate mechanistic relationship. A QSP model extends the analysis by allowing downstream physiological feedback to modify the final response.
14 · Feedback and adaptation

14. Why the Direct Drug Effect Is Not Always the Final Effect

One of the central reasons to use QSP is that biological systems contain feedback.

Suppose a drug directly decreases vascular resistance. The immediate effect may be a fall in arterial pressure. The cardiovascular system can then respond through homeostatic mechanisms.

StepMechanistic event
1Drug concentration increases.
2Drug reduces vascular resistance.
3Mean arterial pressure decreases.
4Baroreceptor signaling changes.
5Autonomic activity changes heart rate and contractility.
6Cardiac output and vascular tone change.
7The final pressure response reflects the combined direct and compensatory effects.

This is an example of why a mechanistic QSP model can provide information that would be difficult to obtain from a simple concentration-versus-blood-pressure regression.

15 · Dynamic behavior

15. Acute Responses Versus Steady-State Responses

Hemodynamic systems can respond on multiple time scales.

Time scaleExamplesPotential QSP processes
Seconds to minutesHeart rate, vascular toneAutonomic responses and direct vascular effects
Minutes to hoursCardiac filling, redistribution of blood flowChanges in venous return and vascular compartments
Hours to daysFluid balanceRenal sodium and water handling
Days to weeksPhysiological adaptationLonger-term neurohormonal and tissue responses

A model that only represents an acute response may be insufficient for questions about chronic treatment. Conversely, a highly detailed long-term model may be unnecessary when the scientific question concerns an immediate pharmacologic response.

Model scope should therefore be determined by the time scale of the scientific question.

16 · Variability

16. Representing Interindividual Variability

Patients do not all have the same cardiovascular physiology. QSP models can represent variability in baseline parameters and physiological responses.

For example, systemic vascular resistance might vary among individuals:

$$SVR_i=SVR_{pop}\exp(\eta_i)$$

where \(SVR_{pop}\) is a typical population value and \(\eta_i\) represents an individual-specific deviation.

Similar variability can be introduced for:

  • Cardiac output.
  • Heart rate.
  • Vascular compliance.
  • Drug sensitivity.
  • Renal fluid handling.
  • Baroreflex sensitivity.

Covariates can then be incorporated when scientifically justified. Age, body size, disease state, baseline cardiovascular function, and concomitant treatment may influence model parameters.

Population QSP perspective: variability is not merely statistical noise. It can provide a mechanistic explanation for why individuals exposed to similar drug concentrations may experience different cardiovascular responses.
17 · Disease systems

17. Extending Hemodynamic Models to Disease

A major advantage of QSP is the ability to represent disease as a perturbation of physiological mechanisms rather than simply as a categorical label.

For example, a cardiovascular disease model might represent changes in:

  • Baseline vascular resistance.
  • Arterial compliance.
  • Cardiac contractility.
  • Preload and venous return.
  • Renal sodium and water handling.
  • Autonomic regulation.
  • Neurohormonal signaling.

The disease state can then be represented by altered parameter values, additional mechanisms, or both.

This creates a mechanistic bridge between normal physiology, disease physiology, and pharmacologic intervention:

$$\text{Normal physiology}\rightarrow\text{Disease perturbation}\rightarrow\text{Drug intervention}\rightarrow\text{Predicted response}$$

Such a framework can be useful for understanding why the same pharmacologic mechanism may produce different responses in different disease states.

18 · Combination therapy

18. Modeling Multiple Drugs

Cardiovascular treatments often act through different mechanisms. A QSP model can represent these mechanisms separately and allow them to interact through the shared physiological system.

Drug mechanismPotential system targetPossible downstream effect
VasodilationVascular resistanceReduced pressure load
Positive inotropyContractilityAltered stroke volume and cardiac output
Chronotropic effectHeart rateAltered cardiac output
Diuresis / natriuresisFluid balanceReduced circulating volume
Neurohormonal modulationRegulatory pathwaysChanges in vascular tone and fluid balance

Because the drugs act on a shared physiological system, their combined response does not necessarily equal the sum of their individual effects. Feedback and nonlinear relationships can produce interactions that emerge only when the complete system is simulated.

19 · Simulation

19. What Can a Hemodynamic QSP Model Simulate?

Once a model has been calibrated and evaluated, simulations can be used to explore scenarios that may be difficult or impractical to test experimentally.

  • Changes in blood pressure following a dose.
  • Heart-rate responses to changes in arterial pressure.
  • Effects of altered vascular resistance.
  • Effects of changes in cardiac contractility.
  • Responses to different dosing regimens.
  • Combination-treatment scenarios.
  • Differences between physiological and disease states.
  • Potential effects of changes in renal fluid handling.
  • Mechanistic explanations for interindividual variability.

For example, a virtual population can be simulated by sampling physiological and pharmacological parameters from distributions representing plausible patient variability.

The resulting simulations can then be summarized as predicted distributions of blood pressure, heart rate, cardiac output, or other endpoints rather than as a single deterministic trajectory.

20 · From biology to model

20. How Is a Hemodynamic QSP Model Built?

Developing a useful QSP model requires translating biological knowledge into a mathematically coherent system.

  1. Define the scientific question. Determine what cardiovascular behavior the model needs to explain or predict.
  2. Define the physiological scope. Decide which cardiovascular, renal, neurohormonal, and pharmacologic processes are necessary.
  3. Define model states and parameters. Identify quantities that change dynamically and parameters that determine their behavior.
  4. Write mechanistic relationships. Convert physiological assumptions into differential and algebraic equations.
  5. Connect pharmacology to physiology. Specify how drug concentration or target engagement modifies physiological processes.
  6. Calibrate the model. Estimate uncertain parameters using appropriate experimental and clinical data.
  7. Evaluate model behavior. Compare predictions with observed data and assess whether the model reproduces important physiological features.
  8. Perform sensitivity analysis. Determine which parameters and mechanisms most strongly influence the outputs of interest.
  9. Simulate scenarios. Use the model to investigate doses, disease states, combinations, or physiological perturbations.

The objective is not to reproduce every detail of cardiovascular physiology. The objective is to construct a model that contains enough mechanistic structure to answer the intended scientific questions.

21 · Sensitivity

21. Sensitivity Analysis in Hemodynamic QSP

Sensitivity analysis asks how strongly model outputs depend on model parameters or mechanisms.

A local normalized sensitivity can be represented conceptually as:

$$S_{y,p}=\frac{\partial y}{\partial p}\frac{p}{y}$$

where \(y\) is an output such as MAP and \(p\) is a model parameter such as vascular resistance, cardiac contractility, or drug potency.

High sensitivity indicates that uncertainty in the parameter can have a substantial influence on the predicted outcome.

Sensitivity analysis can therefore help identify:

  • Which physiological mechanisms dominate a predicted response.
  • Which parameters require more precise estimation.
  • Which measurements would be most informative in future experiments.
  • Where model uncertainty is likely to have the greatest effect on predictions.
Modeling principle: sensitivity analysis is not only a mathematical exercise. It can help connect model uncertainty to experimental design and mechanistic understanding.
22 · Identifiability

22. Identifiability and the Limits of Hemodynamic Data

A mechanistically rich model can contain more parameters than the available data can reliably determine.

For example, if only blood pressure is measured, it may be difficult to determine whether a change arose primarily from cardiac output, systemic vascular resistance, blood volume, or compensatory feedback.

Additional measurements such as:

  • Heart rate.
  • Cardiac output.
  • Stroke volume.
  • Blood pressure.
  • Renal biomarkers.
  • Hormonal measurements.
  • Drug concentrations.

can provide additional constraints on the model.

This is a fundamental QSP principle: model complexity should be supported by information in the available data.

Important distinction: a model may contain a plausible mechanism without that mechanism being identifiable from a particular dataset. Biological plausibility and parameter identifiability are related but distinct concepts.
23 · Interpretation

23. What Hemodynamic QSP Models Do Not Tell Us Automatically

A mechanistic model can be biologically informed and mathematically sophisticated while still being subject to important limitations.

  • Model structure is a simplification. Even a detailed cardiovascular QSP model represents only selected aspects of physiology.
  • Parameter estimates depend on model assumptions. Changing the structure can change estimated parameter values.
  • A good fit does not establish uniqueness. Multiple mechanisms may sometimes explain the same observed endpoint.
  • Unmeasured mechanisms can remain uncertain. A model may contain processes that are weakly constrained by available data.
  • Predictions are conditional. Extrapolation beyond the conditions used for calibration requires additional justification.
  • Feedback can amplify or attenuate direct drug effects. The final physiological response may differ from the immediate pharmacologic action.
  • Virtual populations are model-based. Simulated variability depends on the assumptions and distributions used to construct the population.
Modeling principle: the value of a hemodynamic QSP model lies in its ability to organize biological knowledge, integrate heterogeneous data, generate mechanistic hypotheses, and make quantitatively testable predictions.
24 · Practical workflow

24. A Practical Hemodynamic QSP Workflow

  1. Start with the pharmacologic question. What mechanism or intervention needs to be understood?
  2. Map the physiological pathway. Identify how the drug could affect cardiovascular function.
  3. Define the minimum necessary system. Include the mechanisms needed to answer the question without unnecessary complexity.
  4. Write the governing equations. Translate assumptions about pressure, flow, resistance, volume, and feedback into mathematical relationships.
  5. Connect drug exposure to mechanism. Use an appropriate pharmacodynamic relationship.
  6. Calibrate against multiple data types. Where available, integrate drug concentration, blood pressure, heart rate, cardiac output, and other measurements.
  7. Evaluate model behavior. Check both quantitative fit and qualitative physiological behavior.
  8. Perform sensitivity and identifiability analyses. Determine which parameters and mechanisms are supported by the available information.
  9. Simulate clinically relevant scenarios. Explore doses, disease states, combinations, and patient variability.
  10. Use predictions to generate testable hypotheses. QSP is most useful when model predictions can inform subsequent experiments or clinical development decisions.

25. Key Takeaways

  • Hemodynamics describes the relationships among blood flow, pressure, vascular resistance, cardiac function, and blood volume.
  • QSP models represent these variables as an interconnected physiological system rather than isolated clinical endpoints.
  • Cardiac output is determined by heart rate and stroke volume: \(CO=HR\times SV\).
  • Mean arterial pressure is closely related to cardiac output and systemic vascular resistance, with downstream venous pressure included when appropriate.
  • Vascular resistance provides a mechanistic link between vascular tone and blood flow.
  • Blood volume connects cardiovascular physiology with renal and fluid-balance mechanisms.
  • The baroreflex illustrates how physiological feedback can modify the direct effect of a drug.
  • Drug effects can enter a QSP model by modifying vascular resistance, heart rate, contractility, fluid balance, neurohormonal pathways, or other mechanisms.
  • QSP models can explain why the final physiological response may differ from the immediate pharmacologic effect.
  • Mechanistic cardiovascular models can incorporate disease states, combination therapy, and interindividual variability.
  • Sensitivity analysis helps identify mechanisms and parameters that most strongly influence predicted outcomes.
  • Identifiability depends on the information contained in the available data; a biologically plausible mechanism is not necessarily identifiable from every dataset.
  • The appropriate model is not necessarily the most detailed model. It is the model with sufficient mechanistic structure to address the scientific question and available data.
Next step

Where to Go Next

A natural progression from hemodynamic QSP is to study QSP models of cardiovascular disease, where blood pressure and flow are connected to disease mechanisms such as altered vascular resistance, cardiac dysfunction, fluid retention, and neurohormonal activation.

From there, increasingly integrated models can connect hemodynamics with renal physiology, glucose-insulin regulation, inflammation, pharmacokinetics, pharmacodynamics, and disease progression.

The next tutorial can build directly on the framework introduced here by examining how a cardiovascular QSP model represents cardiac function, vascular resistance, blood volume, and baroreflex regulation together.

← Back to QSP & Pharmacometrics Tutorials