Individualized Therapy of HHT Driven by Network

8/2/2019 Individualized Therapy of HHT Driven by Network

1/24

Individualized therapy of HHTdriven by network

analysis of metabolomic

profiles

Presented by:

Samridhi (2k11/BIO/22)

Shriram (2k11/BIO/)


2/24

Background

Complex diseases with multi-factorialetiologies often have multiple alternative

pathways leading to a particular

pathophysiological state, with a wide range ofresulting phenotypes.

Such diseases provide a significant diagnosticand treatment challenge, and will require

individual-specific, personalized treatments.


3/24

Hereditary Hemorrhagic Telangiectasia (HHT) isan example of a Mendelian genetic disease withbroad variability in presentation and involvementof different organs.

Arteriovenous malformations (AVM) can occur in

multiple beds, including the brain, liver, lungs,and nose.

Since many diseases affect metabolism directly orindirectly, the field of metabolomics has richpotential for biomarker applications.


4/24

A challenge however is that a metabolomic profilealone may not provide any direction for treatment,since there is not biologically coherent integration ofthese data.

Metabolic network reconstructions can provide theframework for integration and analysis of these data.

considered the treatment of an individual patient andthrough comparative analysis of her metabolomic

profile with non-HHT individuals, differences wereidentified using constraint-based analysisof a globalhuman metabolic network reconstruction, Recon 1.


5/24

The patient in this study is a female who beganexperiencing syncopal episodes at rest and duringexertion at 21 years of age.

Further investigation with contrastechocardiography showed she had a large right toleft shunt at the pulmonary capillary level but had

no treatable AVMs on computed tomography (CT)scan of the chest.

Based upon these findings as well as a family

history of familial adenomatous polyposis and asibling with similar diffuse AVMs, a presumeddiagnosis of HHT was given.


6/24

In this individual case study, untargeted,quantitative plasma metabolomic profiling iscarried out in a patient and healthy controls.

used constraint-based modeling of metabolismon an organism scale to identify potentialdifferences in metabolism between the non-HHT

and HHT patient through identification ofbiomarker signature profiles that can be linked todifferent functional states.

These differences were then used to support theuse of particular drug treatments.


7/24

Methods Fasting plasma samples drawn from 5 healthy individuals

(no chronic medical conditions, no current dailymedications or herbal supplements; ages 21-37)

1 patient with HHT (age 24)

Blood draws from two different time points had beenpreviously obtained from two of the healthy individuals andfor the patient, resulting in a total of 9 samples.

Additionally one individual provided a fasting as well asnon-fasting blood sample, which served as a non-fastingpositive control.

Metabolites were identified and quantified


8/24

Sample analysis

Unsupervised hierarchical clustering showedclear separation between the healthy

individuals and the patient.

Separation of the pre-treatment and post-

treatment of the HHT patient as well as

regrouping of the post-treatment HHT patientwith the non-HHT individuals was observed

with Principal Component Analysis as well.


9/24


10/24

Differences in metabolism between non-HHTindividuals and the HHT patient were assessedusing the plasma metabolomic profiles.

Specific metabolites that were quantitatively andqualitatively different between the non-HHT and

HHT patient were determined by identifyingthose metabolites whose maximumconcentration in one group was less than theminimum of the other group or vice versa.

This resulted in two sets of metabolites that weredifferent in the two conditions; a high set and alow set.


11/24

Metabolomic profile


12/24

The two general classes of metabolites werelabeled the as ketone group and the amino

acid/carbon rich group and used to define

qualitative metabolic pseudo-reactions.


13/24

Network analysis Recon 1, a global human metabolic network

reconstruction with elementally charge and massbalanced equations can be used for analysis oftranscriptomic,proteomic, and metabolomic data usingconstraint based analysis methods.

This approach is on the assumptions that following 8hour fasting overnight, the body is at or movingtowards a homeostatic state, the full content of Recon

1 represents the set of metabolic interactions in thehuman body, and that changes in plasma profile reflectnet changes in uptake and/or secretion of differentmetabolites with the set of all organs.


14/24

constraint-based analysis approach for FluxBalance Analysis:

S v = 0

In which S is the mxn stoichiometric matrix withm metabolites, n reactions, with each columnrepresenting a metabolic (or transport) reactionand v is a vector of fluxes corresponding to eachreaction in the network.

Constraints area applied to the network as upperand lower bounds on the fluxes,

v


15/24

The next step requires specification of anobjective function which will then be minimizedor maximized.

linear objective function are considered only inthis study thus,

max(cT v) or min(cT v) are the objectives,in which c is a signed, binary vector.

Flux Variability Analysis (FVA) is a method in

which every reaction in a network is maximizedand minimized under a specified set ofconstraints.


16/24

This approach can be useful when one is interested inproviding a general characterization of the fluxes in onestate and how the extrema change from one conditionto another.

This method has demonstrated interesting results inthe analysis of human metabolism in the general as

well as context specific conditions.

The intravascular space is available for uptake andsecretion of metabolites with all organs of the body,

thus changes in metabolomic profiles may beinterpreted in the context of a global human metabolicnetwork.


17/24

two qualitatively different profiles, derived from

the quantitative differences in the metabolomic

profiles of plasma were used to define two sets ofdifferent transport constraints.

The set of metabolites that were elevated in theHHT patient (i.e. whose minimum measured

concentrations in the HHT patient were greater

than the maximum measured concentrations of

all of the non-HHTs) largely consisted of ketonesand were dubbed the ketone group.


18/24

Conversely, the set of metabolites that weredecreased in the HHT patient (i.e. whosemaximum measured concentrations in the HHT

patient were less than the minimum measuredconcentrations of all of the non-HHTs) primarilywere amino acids and were dubbed the aminoacid group.

The production maxima for the ketone groupand aminoacid group metabolites were used tospecify coefficients for two different pseudo-

objective reaction constraints one representingthe HHT patient and the other representing thenon-HHT profiles.


19/24


20/24

The amino acid group included: tyrosine,taurine.serine, glycine, alanine,citrate, lactate,methanol, and creatinine.

The ketone group included acetone, formate, andacetate.

Subsequently the Flux Span (difference betweenmaximum and minimum flux attainablefor allreactions in a particular condition)

And the Flux Span Ratio (the reaction-wise ratioof the Flux Span between two conditions) werecalculated for comparison

Results


21/24

Results

Comparison of the non-HHT and HHT patient networksusing flux span ratios demonstrated decreased energy

production in the HHT patient, reflective of astarvation-like state.

There were noted increased flux potentials in nitrogen

handling and disposition pathways in the HHTpatient,notably with nitric oxide synthase (NOS).

This observation brought to light a potential link

between vascular endothelial function and the changesin vascularity found in HHT, with connections tometabolism.


22/24

There has been evidence suggesting that VEGF

can decrease blood pressure through

increased nitric oxide production andconversely, that inhibition of VEGF can

increase blood pressure, at least in part,

through the same mechanism.

The potential increase in NOS activity based

on the network analysis of the metabolicprofile supported the use ofbevacizumab

(Avastin), an anti-VEGF drug.


23/24

The HHT patient underwent treatment withbevacizumab at a dose of 5 mg/kg every 2 weeksfor 6 total infusions, at which point another

fasting blood draw was obtained.

The patient had a mild response to therapy whichlasted for 2 months. Surprisingly, the patientsprofile had changed to become more similar tothe control individuals.

When clustered with the other samples, theposttreatment profile clustered with the rest ofthe non-HHT individuals


24/24

Conclusions An individualized case-study is described for a rare disease

in which it is demonstrated how it is possible to progressfrom untargeted metabolomic profiling to identifymetabolic profiles that can be then used to constrain amechanistic network model which is then used to directtherapy decisions.

Metabolomic data has been recognized as an importantomic data type, as it represents a quantitative biochemicalphenotype and has potential to serve as a source ofdiagnostic and therapeutic biomarkers.

In particular, in order to achieve any practical realization ofindividual specific, personalized therapies, it will beimportant to move away from strictly statistically drivenmodels towards more mechanistic models.

Documents

Individualized Therapy of HHT Driven by Network