Métabolisme des phosphates en dialyse

Pr Laurent Juillard,

Dr Sandrine Lemoine

Nephrology department, Lyon.

Inserm CARMEN 1060, Univ Lyon1

CERMEP, Lyon

Conflicts of interest

Clinical research: Alexion, Bayer, GSK, Otsuka

Advisory Boards : Amgen, Baxter, Fresenius, GE Healthcare, Hemotech, Roche, Vifor,

Lectures : Bbraun, Genzyme, GE Heathcare, Vifor, Roche

Research support : Baxter, Fresenius, Genzyme

Compliance issues

All …..

Oral intake

900-1200 mg

490 mg/ 24h

350 mg /24h

Bone = 600-700g

Cells

Porg = 57g

Pi = 3g

Kidney

Interstitial area

Plasma

85%

14%

700mg

Phosphate flux between compartments

Oral intake

900-1200 mg

490 mg/ 24h

350 mg /24h

Bone = 600-700g

Cells

Porg = 57g

Pi = 3g

Kidney

Interstitial area

Plasma

85%

14%

700mg

Phosphate flux between compartments

Plasma Pi :

< 1 % overall phosphorus

= 90 mg

Oral intake

900-1200 mg

490 mg/ 24h

350 mg /24h

Bone = 600-700g

Cells

Porg = 57g

Pi = 3g

Kidney

Interstitial area

Plasma

85%

14%

700mg

Phosphate flux between compartments

Plasma Pi:

< 1 % overall phosphorus

= 90 mg

During dialysis:

= 600 to 700 mg removed

Oral intake

900-1200 mg

490 mg/ 24h

350 mg /24h

Bone = 600-700g

Cells

Porg = 57g

Pi = 3g

Kidney

Interstitial area

Plasma

85%

14%

700mg

Phosphate flux between compartments

Plasma:

< 1 % overall phosphorus

= 90 mg

During dialysis:

= 600 to 700 mg removed

Where does the depurated (Pi) come from ?

Spalding et al, Kidney Int 2002

Time, minutes

Dialysis end

Evolution of plasma Pi

Goals of Pi modeling during dialysis

- Prediction of plasma Pi concentration after changes in dialysis prescriptions (duration, frequency)

Pragmatic, urea type 2 compartments modeling

Description of Pi transfers during dialysis with multicompartimental modeling

TWO POOLS MODELING

Spalding et al, Kidney Int 2002

Evolution of plasma Pi

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 2 pools

2/3 TBW

1/3 TBW

350 mL/min

Spalding et al, Kidney Int 2002

Evolution of plasma Urea

Excellent Fit

Short Long

Spalding et al, Kidney Int 2002

Evolution of plasma bicarbonates

No plateau

Spalding et al, Kidney Int 2002

Evolution of plasma Pi

2 pools modeling

No plateau prediction

Underestimate final Pi

Spalding et al, Kidney Int 2002

Sugisaki et al,Trans Am Soc Artif Intern Organs 1982

Heaf et al, Scand J Urol Nephrol 1998

Evolution of plasma Pi

Spalding conclusion :

Pi prediction impossible with a 2 pools modeling

Confirm Sugisaki and Heaf

AJUSTED TWO POOLS MODELING

Leypoldt et al, Kidney Int 2013

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 2 pools

Distal volume infinite Fixed Km

Infinite pool volume

1/3 TBW

Fixed patient specific Km

Leypoldt et al, Kidney Int 2013

Effet Km

Evolution of plasma Pi 2 pools

Distal volume infinite Fixed Km

Leypoldt et al, Nephrol Dial Transplant 2014

Evolution of plasma Pi 2 pools

Distal volume infinite Fixed Km

Long noctural

Predialysis P

Short daily

Leypoldt et al, Nephrol Dial Transplant 2014

Long noctural

Predialysis P (%)

Short daily

Evolution of plasma Pi 2 pools

Distal volume infinite Fixed Km

Leypoldt et al, Nephrol Dial Transplant 2014

Evolution of plasma Pi 2 pools

Distal volume infinite Fixed Km

Predialysis P (%) 4, 5, 6 Jours

Spalding et al, Kidney Int 2002

Daugerdas et al, Nephrol Dial Transplant 2016

Evolution of plasma Pi 2 pools Kc Variable

3 TBW

1/3 TBW

Kc variable

0,97 m/L

Kc Vaviable selon Pi EC

Daugerdas et al, Nephrol Dial Transplant 2016

Evolution of plasma Pi 2 pools Kc Variable

BLOOD

DIALYSATE

Daugerdas et al, Nephrol Dial Transplant 2016

Evolution of plasma Pi 2 pools Kc Variable

Daugerdas et al, Nephrol Dial Transplant 2016

Evolution of plasma Pi 2 pools Kc Variable

Daugerdas et al, Nephrol Dial Transplant 2016

Evolution of plasma Pi 2 pools Kc Variable

Pi reduction

ratio Mean % error

Training set

Data

Training set

Data

Daugerdas et al, Nephrol Dial Transplant 2016

Evolution of plasma Pi 2 pools Kc Variable

Correct

prediction

Training set

Daugerdas et al, Nephrol Dial Transplant 2016

Evolution of plasma Pi 2 pools Kc Variable

Correct

prediction

Data

COMPLEX COMPARTMENTAL MODELING

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 2 pools

2/3 TBW

1/3 TBW

350 mL/min

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 3 pools

No plateau

Constant gain

Similar parameters

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 3 pools

Constant gain if Pi decreases

Reduced hysteresis

BLOOD CELL

DIALYSATE

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 4 pools

Spalding et al, Kidney Int 2002

Dialysis end

Evolution of plasma Pi 4 pools

Time, minutes

Constant gain if Pi decreases

Pulses from 4th pool

Plateau predicted

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 4 pools

4 pools only predicts adequately

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 4 pools

4 pools allows predictions of treatment change

MR SPECTROSCOPY FOR INTRACELLULAR PI IN HD PATIENTS

BLOOD CELL

DIALYSATE

Hypothesis

EC Pi = IC Pi

Spalding et al, Kidney Int 2002

Evolution of plasma Pi 4 pools

BLOOD CELL

DIALYSATE

Study of Pi transfers is restricted by the lack of direct intracellular Pi measurement

Rational

IC Pi ??

Spalding et al, Kidney Int 2002

PCr = phosphocreatine

α β

γ

ATP

PCr

Pi

Nuclear Magnetic Resonance Spectroscopy Spectrum of 31 phosphorus

- Several peaks like inorganic phosphate (Pi) or phosphocreatin.

- Intracellular concentration assessment of each molecule by repeating acquisitions.

-- Non invasive, non radiant, usable in animals as well as in humans.

Spectrum of 31 phosphorus by NMR spectroscopy

Materials and methods : How to measure intracellular Pi/ IP?

Feasibility of intracellular Pi measurement with spectroRMN during hemodialysis.

Measurement of intracellular Pi and ATP concentration during a hemodialysis session.

Test Spalding’s hypothesis on the diffuse transfert of Pi during hypothesis

Aims of this study

Animal model

•6 anephric pigs, 3 months old, about 30kg.

• Intubated, ventilated, spectroRMN during HD

•Because of the magnetic environment, we performed an adaptation of:

•Blood lines

•Ventilator lines

•Diet = for end stage renal disease, low in potassium, water intake control (500ml/day).

•Dialysis modality:

- Haemodialysis using a PrismaFlex® dialyser.

- Blood flow 100 to 150 mL/min

- Dialysate flow 100 mL/min.

Dialysis modality

Design

Day 1 Day 2 Day 4

Bilateral nephrectomy

HD catheter

Dialysis

31P spectroscopy

Measurements of intracellular and plasma Pi

3H

31P spectroscopy

3H Surgery

0 2000 4000 6000 80000.0

0.2

0.4

0.6

0.8

1.0

Time (s)

Pi/P

CR

Sham pig

Results: Pi kinetic

Times (s)

Pi/P

cr High reproducibility

Lemoine et al, unpublished

Results

Lemoine et al, J Am Soc Nephrol 2016

Results : urea and bicarbonates kinetic

Lemoine et al, J Am Soc Nephrol 2016

Expected evolution of Urea and Bicarbonates

Lemoine et al, J Am Soc Nephrol 2016

Stable calcemia

Calcium balance almost null

Results: Calcium kinetic

Results: Extra and Intracellular Pi kinetic

Lemoine et al, J Am Soc Nephrol 2016

Expected evolution of EC Pi

Unexpected evolution of IC Pi

Constant

depuration of Pi

Results: Intracellular ATP and pH kinetic

Lemoine et al, J Am Soc Nephrol 2016

Unexpected evolution of ATP

Discussion

1. Validation and safety of the model.

2. Opportunity to measure Pi during hemodialysis.

3. Unexpected results :

a. Pi increased in the intra-cellular compartment.

b. ATP decreases

4. Pi origin : intracellular / Bone ?

5. Acute model : to be confirmed in chronic HD patients

BLOOD CELL

DIALYSATE

Adaptation to Spalding 4 pools model

No

Spalding et al, Kidney Int 2002

Yes

EC Pi sensor ??

MR spectroscopy in HD patients

Durozard et al, Kidney Int 1993

MR spectroscopy 4,5 T

Muscle metabolism in 6 normal and HD patients

Day without dialysis

BLOOD

DIALYSATE

Exercise Pcr drop

Pi increase

Durozard et al, Kidney Int 1993

MR spectroscopy in HD patients

Durozard et al, Kidney Int 1993

MR spectroscopy in HD patients

HD

HD

Reduced work load in HD patients

BLOOD

DIALYSATE

EC Pi = IC Pi

Durozard et al, Kidney Int 1993

MR spectroscopy in HD patients

BLOOD

DIALYSATE

Durozard et al, Kidney Int 1993

MR spectroscopy in HD patients

HD

Reduced work load in HD patients

HD

CIPHEMO

12 patients

Dialyse conventionnelle 4 heures

Mesure Pi IC par spectro MR en continu IRM 3T

Evaluation métabolisme osseux

Conclusions

During a hemodialysis:

- Extra-cellular Pi decreased with a plateau as described previously.

- Intra-cellular Pi increased

- Origin of production

- Intra-cellular ATP decreased

- Cell impact ? Fatigue ?

- Spalding model rejected

- Adapted 2 pools modeling may predict Pi after prescription changes

Many thanks to :

Sandrine Lemoine

Amélie Belloi

Thomas Fournier

CERMEP: Gabriel Kocevar, Danielle Ibarola, Dominique Sappey Marinier

HOSPAL