Measuring DLco: What Could Possibly Go Worng?Brian Graham, PhDDivision of Respirology, Critical Care and Sleep MedicineUniversity of Saskatchewan Saskatoon, Canada

Disclosures:In the past 2 years, funding/support has been received from:

• AstraZeneca• Boehringer-Ingelheim• COSMED• GSK• Merck Frosst• Novartis• Nycomed

• Pfizer • Prairie Oxygen• Roche• Trudell• Vitalaire• zu.com

Objectives:

• Describe limitations and sources of error in DLco measurements

• Describe steps to ameliorate their effects

• Describe how current technology can be better applied to DLco measurements

• It’s not just diffusion

• It’s not a capacity

• It’s more precisely called Transfer Factor

Diffusing Capacity of the lung for carbon monoxide: DLco

[single breath]

gas

Palv

PcapDM

flow = pressure x conductance

for CO, assume Pcap = 0

VA·ΔFAco/Δt = PAco · DLco

Krogh equation:DLCO = VA·ln[FACO(t2)/FACO(t1)] PB·(t2-t1)only valid during breath hold

Krogh M. The diffusion of gases through the lungs of man. J Physiol (London) 1914

Roughton & Forster, J Apple Physiol 1957

co

Vc

1. DLco varies with blood volume

2. DLco varies with oxygen tension

1 1 1 . DLco DMco co·Vc

= +

co as PaO2

gasconcen-tration

0

0.3%

time

volume

RV

TLC

0

Ogilvie CM et al. J Clin Invest 1957

breath hold time

DLCO = VA·ln[FACO/FICO ·FITr/FATr] PB·tBH

alveolargas

sample

Ogilvie

gasconcen-tration

0

0.3%

time

volume

RV

TLC

0

breath hold time alveolargas

sample(85ml)

Jones RS, Meade F. Q J Exp Physiol 1961

DLCO = VA·ln[FACO/FICO ·FITr/FATr] PB·tBH

Jones & Meade

gasconcen-tration

0

0.3%

time

volume

RV

TLC

0

Graham et al, IEEE Trans Biomed Eng 1980

VMAX·dFACO(t)/dt = -DLCO·PB·FACO(t)

3 equation

V(t)·dFACO(t)/dt = -DLCO·PB·FACO(t)+(FICO(t)-FACO(t))dV(t)/dt

alveolargas

V(t)·dFACO(t)/dt = -DLCO·PB·FACO(t)

All methods yield similar DLco values in young, healthy, trained subjects

3 equationIdeal J&MOgilvieATS*

Graham et al, J Appl Physiol 1981

fast 10s BH

slow 10s BH

slow 5s BH

150%

125%

100%

0%

normalisedDLco % fast

3 equation

Ideal J&M

Ogilvie

ATS*

Effect of lower flow and shorter breath hold

Graham et al, J Appl Physiol 1981

gasconcen-tration

0

0.3%

time

volume

RV

TLC

0

0.75 litre1.0 litre

ATS/ERS dead space washout - 0.75 – 1 litre - 0.5 litres if VC < 2 litres

ATS/ERS DLco Standardisation. Eur Resp J 2005

Problems with dead space washout

ATS/ERS dead space washout allow 0.75 to 1 litre

1.0 litre0.75 litre

1 litre samplecollection

With current technology we can measure the point of dead space washout so that dead space gas is not included in the alveolar sample

Graham et al. Can Resp J 1996

TLC

Measure dead space washout

Effect of delaying alveolar sample collection

Graham et al, J Appl Physiol 1981

xx x x x x xo

○ Jones & Meade in 10 healthy subjects

x - 3 equation method

- J&M lung model

fast manoeuvre

Effect of delaying alveolar sample collection

Graham et al, J Appl Physiol 1981

xx x x x x xo

x - 3 equation method

- J&M lung model

slow manoeuvre

○ Jones & Meade in 10 healthy subjects

0

gasconcen-tration

0.3%

Calculation of alveolar volume

ATS/ERS DLco Standardisation. Eur Resp J 2005

VI·FITr = VA·FATr + VD·FITr

VA = (VI – VD) · (FITr/FATr)

Measure all of the tracer gas that went onto the lung and subtract all of the tracer gas that came out to get the amount of tracer gas left in the lung at end expiration.Use the concentration measured at end expiration to estimate the concentration of tracer gas in the lung:Lung volume = amount of tracer gas - dead space concentration of tracer gas

Graham et al. Can Resp J 1996

Why not use all of the available information for a more accurate measurement of VA???

VA = FTr(t)dV(t) / FTr(tf) - VDt0

tf

At end exhalation:

0

0.3%

gasconcen-tration

time

volume

RV

TLC

0

Calculate absolute lung volume at end exhalation

CH4 concentration at end exhalation

TLClitres

2 sec 5 sec 10 sec bodybox

breathhold time

Comparison of lung volume measurements

15 normal, healthy trained subjects

ATS3-equation

Graham et al, J Appl Physiol 1985

COPD Patient

CH4CO

pressure

volume

flow

COPD Patient

CH4

CO

TLClitres

2 sec 5 sec 10 sec bodybox

breathhold time

Comparison of lung volume measurements

10 COPD patients

* * *

Graham et al, J Appl Physiol 1985

ATS3-equation

TLClitres

2 sec 10 sec bodybox

breathhold time

Comparison of lung volume measurements

11 Asthma patients

Graham et al, J Appl Physiol 1985

ATS3-equation

• DLco = VA · Kco

• The Krogh coefficient: Kco = ln(Δ[CO])/Δt/PB

• Kco units are min-1 · mmHg-1

• Using ml · min-1 · mmHg-1 · litre is misleading

• Kco should never be labelled DLco/VA

• The relationship between DLco and lung volume is not linear, so DLco/VA and DLco/TLC do NOT provide an appropriate way to normalise DLco for lung volume

Do we really need to measure VA?

ATS/ERS DLco Standardisation. Eur Resp J 2005

YES!

ATS/ERS 2005 Guidelines state:VD can be estimated from various formulae or simply use 150 ml as a fixed value

Effect of VD error on DLco depends on VA, ranging from 1% to 3%

BUT equipment dead space (including filter – up to 350 ml) must also be subtracted or the error can be up to 10%

ATS/ERS DLco Standardisation. Eur Resp J 2005

Calculation of VA: dead space

Graham et al. Can Resp J 1996

TLC

Why not measure VD???

DLco predicted adjustment for Hb

Hb adjustment

males females20% high 175.2 160.8 7.37%10% high 160.6 147.4 3.89%standard 146 134 0.00%10% low 131.4 120.6 -4.38%20% low 116.8 107.2 -9.33%

ATS/ERS DLco Standardisation. Eur Resp J 2005

DLco adjustment: 1.7·Hb /(102.2+Hb) for males

DLco adjustment: 1.7·Hb /(93.8+Hb) for females& children< 15 yrs

• CO in the lung prior to the DLco test has both an “anemia effect” and a CO back pressure effect

• Predicted DLco is reduced 1% for every 2% increase in COHb for COHb levels > 2%

• Baseline COHb in smokers can be as high as 15%• COHb rises with each DLco manoeuvre (almost

1% per manoeuvre in normal healthy subjects)

ATS/ERS DLco Standardisation. Eur Resp J 2005

DLco predicted adjustment for COHb and CO back pressure

DLco predicted adjustment for COHb and CO back pressure

Graham et al. Am J Resp Crit Care Med 2002

fractional concentration of alveolar CO (ppm)

COHb(%)

DLco predicted adjustment for COHb and CO back pressure

Graham et al. Am J Resp Crit Care Med 2002

DLcoml/min/mmHg

nocorrection

FAcocorrection

bothcorrections

controlCOHb = 1.2%

COCOHb = 11.2%

****

Effect of alveolar oxygen on DLco

• DLco increases ~ 0.35% for each 1 mmHg decrease in PAO2

• The alveolar oxygen tension can be calculated from the simplified alveolar gas equation:PAO2 = FIO2(PB-47) - PaCO2 (FIO2 + [1- FIO2] / R)

• As PB decreases or PaCO2 increases,PAO2 decreases and DLco increases

• Correction for PIO2 = 1/(1+0.0031(PIO2 -150))where PIO2 = FIO2(PB-47)

ATS/ERS DLco Standardisation. Eur Resp J 2005

Toronto

Winnipeg

Saskatoon

Edmonton

Calgary

VancouverHalifax Montreal

8%

6%

4%

2%

0%

-2%

Increase in DLCO due to decreasing PIO2 with altitude

Pay attention to the O2 in test gas

• Typically test gas has 21% O2 in North America

• In Europe, many test gas was often 17 - 20%

• Crapo’s DLco reference values were measured with a test gas concentration of 25% to correct for altitude.

• A single breath of test gas at a different O2 might mix sufficiently well in normals but probably not in lung disease patients.

Kendrick Thorax 1993; Crapo Am Rev Resp Dis 1981

Other factors affecting DLco

• Pressure during breath hold: Valsalva Muller

• Volume history – previous deep breath DLco

• Menstrual cycle variations (independent of Hb)

ATS/ERS DLco Standardisation. Eur Resp J 2005

Calibration

• Daily - volume - 3 litre syringe

• Weekly – leak check – use 3 litre syringe as test subject – “biologic control”

• Quarterly – gas analyser linearity check

ATS/ERS DLco Standardisation. Eur Resp J 2005

Summary

• The standardised DLco test was not designed to measure gas exchange in patients with airflow obstruction

• The standardised DLco test remains constrained by limitations to the original instrumentation and manual computation

• Using existing equipment, more accurate and precise DLco measurements can be made from a patient-friendly, more physiologic manoeuvre

Summary

• In spite of the many limitations of the standardised test, DLco is a valuable measure of pulmonary function, but it has the potential to be a much more useful test

If the application of technology to communication had proceeded in the same manner as the application of technology to diffusing capacity, this is what a smart phone would like today

Thanks to the Saskatchewan DLco Team

David J CottonJoseph T Mink