C

PPa

b

c

d

e

f

a

AA

KCERVV

1

uwvrecdrceececb

P3

1d

Respiratory Physiology & Neurobiology 179 (2011) 167– 173

Contents lists available at ScienceDirect

Respiratory Physiology & Neurobiology

j our nal ho me p age: www.elsev ier .com/ locate / resphys io l

ardiopulmonary exercise testing (CPET) in pulmonary emphysema

atrizia Paoletti a , Francesca De Filippisa , Francesco Fraiolib , Alessandra Cinquantac , Gabriele Valli a ,ierantonio Lavenezianaa,d,e, Francesco Vaccaroa, Dario Martolinia, Paolo Palangea,f,∗

Department of Public Health and Infectious Diseases, Pulmonary Lung Function Unit, “La Sapienza” University, Rome, ItalyDepartment of Radiology, “La Sapienza” University, Rome, ItalySiemens Italia, Milan, ItalyLaboratoire de Physio-Pathologie Respiratoire, Equipe de Recherche ER 10 UPMC, Faculté de Médecine Pierre et Marie Curie (site Pitié-Salpêtrière), Université Paris VI, Paris, FranceService de Pneumologie et Réanimation, Groupe Hospitalier Pitié-Salpêtrière, Paris, FranceFondazione Eleonora Lorillard Spencer Cenci, Rome, Italy

r t i c l e i n f o

rticle history:ccepted 29 July 2011

eywords:ardiopulmonary exercise testingmphysemaespiratory mechanicsentilatory efficiency

a b s t r a c t

In patients affected by chronic obstructive pulmonary disease (COPD), cardiopulmonary response toexercise was never related to the severity of emphysema (E) measured by high resolution computedtomography (HRCT).

Sixteen patients (age = 65 ± 8 yrs; FEV1 = 54 ± 18%pred; RV = 160 ± 28%pred) with moderate to severeE (quantified by lung HRCT as % voxels < −910 HU) were exercised on a cycle-ergometer to exhaustion.Oxygen uptake (V̇O2 ), carbon dioxide output (V̇CO2 ), ventilation (V̇E), tidal volume (VT), and end-tidal PCO2

(PETCO2 ) derived variables were measured breath-by-breath.˙ ˙

entilatory response The % of E correlated with: (1) the ratio VTpeak /FEV1 (r = 0.74; p = 0.001); (2) the VE/VCO2 slope(r = −0.77; p = 0.0004); (3) PETCO2 values at peak exercise (r = 0.80; p = 0.0001). Also, the %E was stronglypredicted by the following exercise equation: %EEST = 58.1 + 11.9 × VTpeak /FEV1 − 0.8 × �V̇E/�V̇CO2

(r = 0.94; p < 0.0001).A V /FEV ratio > 1 is typically observed in severe E patients; furthermore, the V̇ /V̇ slope and the

e and

Tpeak 1

PETCO2 peak values decreas

. Introduction

Cardiopulmonary exercise testing (CPET) has been increasinglysed for evaluating the causes of exercise intolerance in patientsith lung and heart diseases (Palange et al., 2007). In particular

entilatory limitation to exercise has been defined as the markededuction, or absence, of breathing reserve (BR) measured at peakxercise (Wasserman et al., 1994) that is typically observed inhronic lung disease states such as chronic obstructive pulmonaryisease (COPD). In COPD, however, airflow obstruction could be theesult of intrinsic airway narrowing, as observed in predominanthronic bronchitis profile, or may be the result of the loss of lunglastic recoil, as observed in patients with prevailing pulmonarymphysema profile. Importantly these anatomical alterations bothause a reduction in expiratory flow reserve. BR is commonly

xpressed as the ratio of ventilation (V̇E) measured at peak exer-ise over the maximal ventilatory capacity (MVC) that is calculatedy multiplying the forced expiratory volume in 1 s (FEV1) by 40

∗ Corresponding author at: Department of Public Health and Infectious Disease,ulmonary Lung Function Unit, University of Rome ‘La Sapienza’, Viale Università7, 00185 Rome, Italy. Tel.: +39 06 49 40 421; fax: +39 06 49 40 421.

E-mail address: paolo.palange@uniroma1.it (P. Palange).

569-9048/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.resp.2011.07.013

E CO2

increase respectively as more as the emphysema is severe.© 2011 Elsevier B.V. All rights reserved.

(Campbell, 1982). However, it is well known that in COPD thedegree of exercise intolerance cannot be adequately predicted byresting FEV1 measurement and the severity of emphysema quanti-fied by high resolution computed tomography scan (HRCT) varieswidely despite the same disease stage (Kurashima et al., 2005;Makita et al., 2007; Omori et al., 2008). Therefore some COPDpatients may show very little evidence of emphysema whereasothers may show marked emphysema at HRCT regardless of theirresting FEV1. Furthermore, some patients retain relatively normalpulmonary function despite the presence of severe emphysema atHRCT.

We hypothesized that in patients with emphysema, because ofthe loss of elastic recoil and early collapse of the airways of mediumcalibre during the forced expiratory manoeuvre, FEV1 values couldoverestimate the severity of airflow limitation. If so, in patients withE, the tidal volume (VT) measured at peak exercise may exceed theresting FEV1 value, being this disproportion as greater as more theemphysema is severe. In addition, to the best of our knowledge, nostudies have examined the CPET variables during maximal incre-mental exercise in patients with emphysema quantified at HRCT.

In particular the slope of the V̇E over CO2 production (�V̇E/�V̇CO2 )and the variation in the end-tidal PCO2 (PETCO2 ) values duringCPET in patients presenting with different severity of emphysemaquantified at HRCT have not yet been described in detail in the

168 P. Paoletti et al. / Respiratory Physiology & Neurobiology 179 (2011) 167– 173

Table 1Baseline characteristics of patients (n = 16) and healthy controls (n = 10).

Patients Controls p-Value

Age, yr 65 ± 8 65 ± 5 NSCigarette smoking, packs/yr 33 ± 21 0.6 ± 1.7 0.0001BMI, kg/m2 25 ± 4 27 ± 3 0.05FEV1, l (%pred) 1.6 ± 0.6 (54 ± 18) 2.7 ± 0.5 (114 ± ± 13) 0.0001FVC, l (%pred) 3.2 ± 0.8 (87 ± 21) 3.6 ± 0.8 (106 ± 12) NSFEV1/FVC, % 48 ± 11 75 ± 4 0.0001MEF50/MIF50 0.21 ± 0.07 0.78 ± 0.15 0.0001TLC, l (%pred) 7.2 ± 1.4 (109 ± 24) 6.2 ± 1.2 (110 ± 7) NSRV, l (%pred) 4.0 ± 1.3 (160 ± 28) 2.3 ± 0.4 (108 ± 14) 0.0005DLCO, ml/min/mmHg 69 ± 14 108 ± 14 <0.0001PaO2 , mmHg 77 ± 9 / /PaCO2 , mmHg 40 ± 3 / /

V ital cafl capacc

ls

2

2

spcecdEcigfriiaHSHt

2

acucRart4eiWac

alues are mean ± SD. FEV1, forced expiratory volume in one second; FVC, forced vow at 50% of FVC; TLC, total lung capacity; RV, residual volume; DLCO, diffusion

arbon dioxide arterial partial pressure. NS, not significant.

iterature. This was, therefore, the second purpose of the currenttudy.

. Methods

.1. Study subjects

Sixteen patients with moderate to severe emphysema weretudied. They were males, aged 65 ± 8 yrs and ex-smokers (33 ± 21acks/yr) (Table 1). Patients with a prior diagnosis of otherardio-respiratory conditions (i.e., asthma, interstitial lung dis-ases, primary pulmonary hypertension, manifest cor pulmonale,hronic heart failure, diabetes, muscular and peripheral vasculariseases) that could affect exercise performance were excluded.xclusion of co-morbidities was done by medical history, physi-al examination, patients’ past medical records, and by additionalnvestigations (laboratory, radiological and so on) if necessary. Aroup of 10 age-matched healthy control subjects was also includedor two reasons: (1) to confirm that the VTpeak

never exceeds theesting FEV1 value in normal subjects and (2) to compare breath-ng pattern, VTpeak

/FEV1 and ventilatory response (�V̇E/�V̇CO2 ) toncremental cycle CPET of healthy subjects with those of patientsffected by emphysema. Healthy subjects were not evaluated byRCT. The study was approved by the ethic committee of “Laapienza” University of Rome in agreement with the declaration ofelkinsi. Written informed consent was obtained from each par-

icipant before the initiation of the study.

.2. Pulmonary function testing

To quantify the lung function impairment in the study groupnd to exclude the presence of a lung function impairment in theontrol group, all participants underwent a full lung function eval-ation including spirometry (COSMED PFT, Rome, Italy), diffusingapacity of the lung for carbon monoxide (DLCO) (COSMED PFT,ome, Italy), body plethysmography (COSMED PFT, Rome, Italy)nd, only in emphysema patients, arterial blood gas analysis atest and at peak exercise in order to assess the arterial CO2 par-ial pressure (PaCO2 ) (Instrumentation Laboratory, GEM premier000), as previously described (Laveneziana et al., 2011; Palanget al., 2004). Baseline spirometry and lung volumes were assessed

n accordance with recommended techniques (Miller et al., 2005;

anger et al., 2005). Measurements were standardized as percent-ges of predicted normal values (Quanjer et al., 1993). The baselineharacteristics of patients and controls are illustrated in Table 1.

pacity; MIF50, maximal inspiratory flow at 50% of FVC; MEF50, maximal expiratoryity of the lung for carbon monoxide; PaO2 , oxygen arterial partial pressure; PaCO2 ,

2.3. Assessment of emphysema

To quantify the severity of emphysema all patients underwent tolung HRCT performed at full inspiration. The images were acquiredwith a 64-detector row spiral CT scanner (Volume Sensation Car-diac, Siemens, Erlangen, Germany). The parameters used for thescan acquisition were as follows: 100 kV, CAREdose with qual-ity reference set at 120 mA s; collimation 64 mm × 0.6 mm; gantryrotation time 0.33 s; scan time 6 s; reconstruction kernel B30 andB60. A dedicated software (Lung Parenchima Analysis Software,Siemens) was used for an automated analysis of the lung air vol-umes and emphysema extent estimation by using a dedicatedsoftware. Emphysema extent (%E) was quantified and expressedas the percentage of the total lung volume having an attenuationranging between −1024 and −910 HU. The −910 HU threshold wasa pre-defined threshold value used as a default setting by the soft-ware and it was applied in recent studies to quantify the severity ofemphysema (Park et al., 1999; Hoffman et al., 2009; Gietema et al.,2010).

2.4. Exercise testing

Both emphysematous patients and controls performed a maxi-mal incremental exercise protocol on a electromagnetically brakedcycle ergometer (Bosch, ERG 551, Germany) consisting of (a) 2 minat rest; (b) 1 min unloaded warm-up cycling; (c) individually-tailored incremental phase (10–15 W/min); (d) 4 min recoveryperiod. Participants were asked to cycle at a pedalling rate of60 rpm; the test was conducted until exhaustion. Oxygen uptake(V̇O2 , STPD), carbon dioxide production (V̇CO2 , STPD), ventila-tion (V̇E, BTPS), tidal volume (VT), respiratory frequency (fR) andend-tidal CO2 partial pressure (PETCO2 ) were measured breath-by-breath (COSMED Quark b2, Rome, Italy) and averaged every 10 s forsubsequent analysis. The cardiac frequency (fH) was monitored via12 leads ECG. Arterial oxygen saturation was monitored by the useof a pulse oxymeter (SpO2). Peak work rate (WR), peak V̇O2 and peakV̇E were defined, respectively, as the highest level of exercise andthe highest V̇O2 and V̇E that could be sustained for at least 15 s duringthe last stage of incremental exercise. Maximal ventilatory capacity(MVC) was estimated as FEV1 × 40 (Campbell, 1982). The anaerobicthreshold (LT) was detected by the V-slope method (Beaver et al.,1986) and supporting ventilatory and pulmonary gas exchangecriteria (Palange et al., 2007). Iso-WR was defined as the highestequivalent exercise WR achieved by all participants during all the

tests. The slope of V̇E over V̇CO2 (�V̇E/�V̇CO2 ) during incrementaltest was measured from unloaded pedalling to the ventilatory com-pensation point (VCP) and, for patients who did not reach the VCP(75% of cases) it was measured from unloaded pedalling to peak

P. Paoletti et al. / Respiratory Physiology &

Table 2Incremental exercise data of patients (n = 16) and healthy controls (n = 10).

Patients Controls p-Value

WRmax, W 88 ± 22 160 ± 37 <0.0001V̇O2 peak l/min 1.57 ± 0.35 2.05 ± 0.62 <0.0001V̇O2 /fH peak, ml/beat 12 ± 2 12 ± 2 NSV̇E peak, l/min 54 ± 11 70 ± 17 0.007V̇E/MVC peak 0.93 ± 0.31 0.60 ± 0.10 0.003VT peak, l 1.74 ± 0.37 2.10 ± 0.79 NSVT peak/FEV1 1.17 ± 0.28 0.76 ± 0.15 0.0003fR peak, breaths/min 33 ± 5 35 ± 5 NS�V̇E/�V̇CO2 peak 32 ± 5 30 ± 4 NS�V̇E/�V̇CO2 29 ± 5 26 ± 3 0.02PETCO2 peak, mmHg 36 ± 4 39 ± 5 0.03

Values are mean ± SD. WR, work rate; V̇O2 , oxygen uptake; fH, cardiac frequency;V̇E, ventilation; MVC, maximal ventilatory capacity; VT, tidal volume; fR, respira-t�p

ewfdr2tw(uce

2

fSuvbpedwp%ocp

3

4Tgt(adpw(

VTpeak/FEV1 ≥ 1 had a %E ≥ 50%.

In 10 (n = 5 with %E ≤ 50% and n = 5 with %E ≥ 50%) out of 16patients, PaCO2 and IC were measured at rest and at peak exercise.The peak-to-rest change in PaCO2 (�PaCO2 ) was significantly higher

Table 3Comparison between groups with different % of emphysema (E) quantified by HRCT.

%E ≤ 50% (n = 6) %E > 50% (n = 10) p-Value

FEV1/FVC, % 58 ± 8 42 ± 8 0.002MEF50/MIF50 0.33 ± 0.12 0.18 ± 0.06 0.005VT peak, l 1.69 ± 0.44 1.77 ± 0.34 NSVT peak/FEV1 0.91 ± 0.13 1.32 ± 0.23 0.001V̇E peak, l/min 53.7 ± 10 54.7 ± 12 NSV̇CO2 peak, l/min 1.55 ± 0.44 1.80 ± 0.44 NS�V̇E/�V̇CO2 33 ± 5 27 ± 3 0.003V̇E/V̇CO2 peak 36 ± 6 31 ± 3 0.03PETCO2 peak, mmHg 32 ± 5 39 ± 2 0.001

Values are mean ± SD. HRCT, high resolution computed tomography; FEV1, forced

ory frequency; V̇CO2 , carbon dioxide production; V̇E/V̇CO2 , the ratio of V̇E over V̇CO2 ;V̇E/�V̇CO2 , the slope of V̇E over V̇CO2 ; PETCO2 , end-tidal carbon dioxide partial

ressure. NS, not significant.

xercise. Two experts independently read each test and the resultsere averaged. Inspiratory capacity (IC) manoeuvres were per-

ormed at rest and at peak-exercise only in emphysema patients:ynamic lung hyperinflation was defined by any decrease in IC fromest to peak exercise, as previously described (Laveneziana et al.,011; O’Donnell et al., 2001). Peak-to-rest changes (�) in IC and inhe ratio (expressed as %) of VT over IC (�VT%IC) during exerciseere used as indices of volume constraints, as described elsewhere

O’Donnell et al., 2001). The physiological dead space-to-tidal vol-me ratio (VD/VT), an index of lung gas exchange efficiency, wasalculated by using the formula (Laveneziana et al., 2011; Palanget al., 2004):

VD

VT= 1 −

(VCO2

VE× 863

PaCO2

)

.5. Statistical analysis

Results are expressed as mean ± standard deviation (SD). Dif-erences between patients and controls data were determined bytudent’s t-test. The normality of data distribution was assessedsing the Shapiro–Wilks test. Relationships between physiologicariables of interest were tested by linear regression analysis andy Pearson regression coefficient (R). Variables which were eitherhysiologically plausible or significantly correlated with %E werentered into multiple linear regression analysis with %E as theependent variable and independent variables selected by a step-ise procedures. In order to analyse possible exercise test variablesredictive of severe emphysema extent, we select two groups withE ≤ 50% and >50%; Student’s t-test was used to compare continu-us variables between two groups. A p value less than 0.05 wasonsidered to be statistically significant. Statistical analysis waserformed by using the SPSS 10.0 software.

. Results

In our patients the %E varied between 36 and 56% (mean value9 ± 6%). Exercise data of patients and controls are presented inable 2. On average exercise tolerance was reduced in the studyroup (V̇O2 peak, 1.57 ± 0.35 l/min (78% predicted). The duration ofhe cycle CPET was 11 ± 2 min, with no significant differencesp = 0.13) between patients presenting with %E ≤ 50% (10 ± 2 min)nd those presenting with %E ≥ 50% (11 ± 2 min). The LT was

etected in 4/6 patients presenting with %E ≤ 50% and in 6/10atients presenting with %E ≥ 50%. At LT no significant differencesere found between the two groups (%E ≤ 50% vs %E ≥ 50%) in V̇O2

0.99 ± 0.10 l/min vs 1.18 ± 0.28 l/min, p = 0.13), V̇O2 %pred (65 ± 2%

Neurobiology 179 (2011) 167– 173 169

vs 66 ± 7%, p = 0.67), V̇CO2 (0.90 ± 0.21 l/min vs 1.05 ± 0.26 l/min,p = 0.35), V̇E (30.9 ± 2.8 l/min vs 33.1 ± 6.3 l/min, p = 0.46), V̇E/V̇O2

ratios (31 ± 2 vs 28 ± 3, p = 0.11), V̇E/V̇CO2 ratios (35 ± 5 vs 32 ± 4,p = 0.35) and PETCO2 (36 ± 5 mmHg vs 38 ± 2 mmHg, p = 0.34). Thegreat majority of our emphysema patients did not reach the VCP. Inthis regard, VCP was detected only in 2/6 patients presenting with%E ≤ 50% and only in 2/10 patients presenting with %E ≥ 50%, thusmaking highly unreliable a comparison of data at VCP between thetwo groups.

Analyzing the breathing pattern during exercise, we foundthat, compared to control subjects in which the VT peak neverexceeded the resting FEV1 value, in 63% of emphysematouspatients (n = 10) the VT peak/FEV1 was greater than 1 varying,for the entire group, between 0.74 and 1.71 (mean 1.17 ± 0.28)(Table 2). Furthermore, VT peak resulted on average 55% of forcedvital capacity (FVC). Mean V̇E peak was significantly lower in patientscompare to controls (54 ± 11 l/min vs 70 ± 17 l/min, p = 0.007)while no differences in mean fR at peak exercise were found(Table 2). Patients stopped exercising because of ventilatorylimitation as suggested by reduced BR at peak exercise (V̇E/MVC = 0.93 ± 0.31) (Table 2). The ventilatory response did notsignificantly differ between patients with VTpeak

/FEV1 > 1 and

those with VTpeak/FEV1 < 1 throughout exercise: V̇E was simi-

lar at rest (15 ± 3 l/min vs 12 ± 2 l/min, respectively), at 20 W(24 ± 5 l/min vs 20 ± 2 l/min, respectively), at 40 W (which repre-sents the iso-WR, 31 ± 5 l/min vs 30 ± 6 l/min, respectively), and atpeak exercise (54 ± 10 l/min vs 53 ± 10 l/min, respectively). The VTexpressed as % of measured slow vital capacity (SVC) did not sig-nificantly differ between patients with VT peak/FEV1 > 1 and thosewith VT peak/FEV1 < 1 at rest (31 ± 12% vs 29 ± 10%, respectively),at 20 W (45 ± 13% vs 37 ± 14%, respectively), at 40 W (51 ± 10% vs47 ± 9%, respectively) and at peak exercise (55 ± 11% vs 53 ± 4%,respectively).

As expected, in emphysematous patients the mean �V̇E/�V̇CO2 was higher compared to healthy subjects (29 ± 5 vs 26 ± 3,p = 0.02) (Table 2). In patients, mean PETCO2 peak value was lowercompared to healthy subjects (36 ± 4 mmHg vs 39 ± 5 mmHg,p = 0.03) (Table 2). The study patients did not present with a sig-nificant oxygen desaturation at the end of exercise.

Dividing the patients in two groups based on the percentageof E (i.e., %E ≤ 50 and %E ≥ 50), we observed a significant dif-ference between them for FEV1/FVC, MEF50/MIF50, VT peak/FEV1,�V̇E/�V̇CO2 and PETCO2 peak (Table 3). The totality of patients with

expiratory volume in 1 s; FVC, forced vital capacity; MEF50, maximal expiratory flowat 50% of FVC; MIF50, maximal inspiratory flow at 50% of FVC; VT, tidal volume; V̇E,ventilation; V̇CO2 , carbon dioxide production; �V̇E/�V̇CO2 , the slope of V̇E over V̇CO2 ;V̇E/V̇CO2 , the ratio of V̇E over V̇CO2 ; PETCO2 , end-tidal carbon dioxide partial pressure.NS, not significant.

170 P. Paoletti et al. / Respiratory Physiology &

Tab

le

4C

omp

aris

on

betw

een

pat

ien

ts

wit

h

%E

≤

50%

(fro

m

#1

to

#5)

and

pat

ien

ts

wit

h

%E

≥

50%

(fro

m

#6

to

#10

)

in

wh

om

arte

rial

bloo

d

gas

sam

pli

ng

and

insp

irat

ory

cap

acit

y

mea

sure

men

ts

wer

e

asse

ssed

.

%E

FEV

1/F

VC

MEF

50/

MIF

50

V̇C

O2

pea

k

(l/m

in)

V̇E

pea

k

(l/m

in)

VT p

eak/

FEV

1�

IC

(l)

VT%

ICp

eak

�V

T%

IC

�V̇

E/

�V̇

CO

2V̇

E/

V̇C

O2

pea

kPE

T CO

2p

eak

(mm

Hg)

PaC

O2

pea

k(m

mH

g)�

PaC

O2

(mm

Hg)

�V

D/V

T

#1

43

62

0.40

1.40

43

0.68

−0.5

0

92

64

26

31

38

49

5

−0.0

7#

2

46

59

0.29

1.15

45

0.97

−0.0

9

79

44

32

39

32

43

−5

−0.1

7#

3

36

53

0.22

1.72

75

0.82

0

81

49

41

44

29

43

−1

−0.2

0#

4

40

52

0.30

1.43

55

0.85

−0.6

5

80

51

32

39

28

43

1 −0

.17

#5

45

52

0.30

1.50

56

0.98

0

76

38

32

37

33

35

−6

−0.3

1

Mea

n

±

SD

42

±

4

56

±

5

0.30

±0.0

6

1.44

±

0.20

55

±

13

0.86

±

0.12

−0.2

5

±

0.31

81

±

6

49

±

10

33

±

5

38

±

5

32

±

4

43

±

5

−1

±

4

−0.1

8

±

0.09

#6

56

42

0.16

2.16

60

1.15

−1.5

0

100

81

23

28

41

46

7

−0.1

6#

7

53

36

0.13

1.93

54

1.40

−0.4

8

100

73

25

28

44

47

9

0.08

#8

52

51

0.28

2.67

67

1.00

−0.8

7

98

72

22

25

40

45

3

−0.3

3#

9

51

41

0.10

1.22

37

1.31

−0.7

7

93

66

26

30

40

42

5

−0.0

6#

10

51

48

0.25

1.95

54

1.11

−1.1

0

90

67

23

28

39

40

2

−0.0

7

Mea

n

±

SD

52

±

2

44

±

6

0.18

±

0.08

1.99

±

0.52

54

±

11

1.19

±

0.16

−0.9

4

±

0.38

96

±

5

72

±

6

24

±

2

28

±

2

41

± 2

44

±

3

5

±

3

−0.1

1

±

0.15

p-V

alu

e

0.00

1

0.01

0.03

0.06

0.9

0.00

6

0.01

0.00

3

0.00

2

0.00

8

0.00

2

0.00

2

0.6

0.03

0.4

Val

ues

are

mea

n

±

SD. %

E,

%

of

emp

hys

ema

quan

tifi

ed

by

hig

h

reso

luti

on

com

pu

ted

tom

ogra

ph

y;

FEV

1,

forc

ed

exp

irat

ory

volu

me

in

one

seco

nd

; FV

C, f

orce

d

vita

l cap

acit

y;

MEF

50, m

axim

al

exp

irat

ory

flow

at

50%

of

FVC

; MIF

50,

max

imal

insp

irat

ory

flow

at

50%

of

FVC

;V̇

CO

2, c

arbo

n

dio

xid

e

pro

du

ctio

n;

V̇E, v

enti

lati

on;

VT, t

idal

volu

me;

�IC

, pea

k-to

-res

t

dif

fere

nce

(�)

in

insp

irat

ory

cap

acit

y (I

C);

VT%

IC, t

he

rati

o

(exp

ress

ed

as

%)

of

VT

over

IC;

�V

T%

IC,

pea

k-to

-res

t

dif

fere

nce

(�)

in

the

rati

o

(exp

ress

ed

as

%)

of

VT

over

IC; �

V̇E/

�V̇

CO

2, t

he

slop

e

ofV̇

Eov

erV̇

CO

2;V̇

E/

V̇C

O2, t

he

rati

o

ofV̇

Eov

erV̇

CO

2; P

ETC

O2, e

nd

-tid

al

carb

on

dio

xid

e

par

tial

pre

ssu

re; P

a CO

2, t

he

arte

rial

carb

on

dio

xid

ep

arti

al

pre

ssu

re;

�Pa

CO

2, p

eak-

to-r

est

dif

fere

nce

(�)

in

PaC

O2;

�V

D/V

T, p

eak-

to-r

est

dif

fere

nce

(�)

in

the

ph

ysio

logi

cal d

ead

spac

e-to

-tid

al

volu

me

rati

o.

NS,

not

sign

ifica

nt.

Neurobiology 179 (2011) 167– 173

in the %E ≥ 50% than in the %E ≤ 50% (5 ± 3 mmHg vs −1 ± 4 mmHg,p = 0.03), clearly meaning that patients with more severe emphy-sema were prone to increase PaCO2 during exercise more thanpatients with less severe emphysema (Table 4). The peak-to-restchange in VD/VT (�VD/VT) did not differ between %E ≤ 50% and%E ≥ 50% (−0.18 ± 0.09 vs −0.11 ± 0.15, p = 0.4) (Table 4). Dynamiclung hyperinflation as defined by any decrease in IC from rest topeak exercise (�IC) was experienced by the majority of patients(8/10). However, the extent of dynamic lung hyperinflation wassignificantly less in %E ≤ 50% than in %E ≥ 50% (−0.25 ± 0.31 l vs−0.94 ± 0.38 l, p = 0.01) (Table 4). In addition, peak-to-rest changein the VT/IC ratio (�VT%IC) during exercise, an index of progres-sive mechanical/volume constraint, was significantly greater in themore severe emphysema group than in the less severe emphysemagroup (72 ± 6 vs 49 ± 10, p = 0.002) (Table 4).

3.1. Correlations

The ratios FEV1/FVC and maximal expiratory flow over maxi-mal inspiratory flow at 50% of FVC (MEF50/MIF50) were the onlyresting functional measurements significantly correlated with %E(r = −0.72; p = 0.002 and r = −0.65; p = 0.006, respectively) (Fig. 1A).V̇O2 peak was not predicted by the severity of emphysema at lungHRCT and by all resting lung function indexes. As expected,the absolute value of VT peak correlated with both FEV1 (r = 0.74;p = 0.001) and FVC (r = 0.78 p = 0.0004). Interestingly, the ratioVT peak/FEV1 directly correlated with %E (r = 0.74; p = 0.001) (Fig. 1B).The analysis of ventilatory response during exercise showed anegative correlation between the %E and both V̇E/V̇CO2 value atpeak exercise (r = −0.72; p = 0.02) and �V̇E/�V̇CO2 during exercise(r = −0.77; p = 0.0004) (Fig. 1C). We found a positive correlationbetween the %E and the value of PETCO2 peak (r = 0.80; p = 0.0001)(Fig. 1D). After performing a multiple regression model with twoindependent variables selected by a stepwise procedure, we foundthat the greater proportion of %E was predicted by the followingequation:

%EEST = 58.1 + 11.9 × VTpeak/FEV1

− 0.8 × �V̇E/�V̇CO2 (r = 0.94; p < 0.0001)

with the linear regression of %EEST on %E having a slope not signif-icantly different from unity (0.85 ± 0.1; p < 0.0001) (Fig. 2).

4. Discussion

The main findings of this study are as follows: (1) during incre-mental exercise patients with emphysema are able to increase theVT relatively to FEV1 as more as greater is the severity of emphy-sema; (2) the levels of V̇E required to clear the CO2 produced duringexercise (i.e., V̇E/V̇CO2 slope) are reduced in patients with moresevere emphysema; (3) VT peak/FEV1 values and V̇E/V̇CO2 slopes areuseful to predict the severity of emphysema detected at lung HRCT.

4.1. VT peak/FEV1 relationship

In COPD airflow obstruction could be the result of intrinsicairway narrowing, as observed in predominant chronic bronchi-tis profile, or may be the result of the loss of lung elastic recoil,as observed in patients with prevailing pulmonary emphysemaprofile. Importantly these anatomical alterations both cause areduction in expiratory flow reserve. In patients with emphysema,

because of the loss of elastic recoil and early collapse of the air-ways of medium calibre during the forced expiratory manoeuvre,FEV1 values could overestimate the severity of airflow limita-tion. The corollary of this is that VT measured at peak exercise

P. Paoletti et al. / Respiratory Physiology & Neurobiology 179 (2011) 167– 173 171

F eak /FE9

mgsVoisoIoeiosVRipeiapf

r

ig. 1. Correlations between %ECT and respiratory variables: FEV1/FVC (panel A), VTp5% confidence interval.

ay exceed the resting FEV1 value, being this disproportion asreater as the emphysema is more severe. The results of the presenttudy confirmed our hypothesis that in patients with emphysemaT peak may exceed the resting FEV1, proportionally to the severityf emphysema detected at lung HRCT (Table 3). Previous stud-es have characterized the breathing pattern response of healthyubjects and COPD patients during a maximal incremental CPETn cycle ergometer (Blackie et al., 1991; O’Donnell et al., 2001).n normal individuals it was reported a mean VT peak/FEV1 ratiof 0.75–0.85 and a mean VT peak/FVC ratio of 0.57–0.63 (Blackiet al., 1991; O’Donnell et al., 2001). In agreement with these results,n our control group the VT peak/FEV1 was 0.76 ± 0.15. Two previ-us studies examined the VT response during incremental CPET inevere COPD patients but did not focus on the difference in theT peak/FEV1 pattern (Blackie et al., 1991; Vogiatzis et al., 2005).ecalculating the mean VT peak/FEV1 measured in these two stud-

es, we obtained a value of 1.17 and 1.31, respectively. In 63% of ouratients VT peak/FEV1 values were greater than 1 and in all patientsxceeded 0.75. More importantly, a VT peak/FEV1 > 1 was almostnvariably associated to greater extent of emphysema (i.e., %E ≥ 50%t lung HRCT). Of note, mean VT peak/FVC values observed in the

resent study did not differ from those reported in the literatureor healthy subjects.

The most likely explanation for the observed result is that, atest, the more advanced the severity of emphysema the greater

V1 (panel B), �V̇E/�V̇CO2 (panel C) and PETCO2 peak (panel D). Dashed lines represent

is the bronchial airways narrowing during the forced expiratorymanoeuvre, and, consequently, the reduction of FEV1 proportion-ally to FVC (Table 3). During exercise the positive pressure exertedon the bronchial walls is less and patients with more severe emphy-sema can achieve a VT value greater than that one expected on thebasis of FEV1. This hypothesis is supported by the evidence of a sig-nificant correlation between %E and both the ratio FEV1/FVC and theratio MEF50/MIF50, with preservation of a relatively normal VT/FVCratio.

Although the small sample size of the current study precludesa definitive conclusion on the ability of VT/FEV1 value obtained atpeak CPET to evaluate the severity of pulmonary emphysema, thedetection of a high VT peak/FEV1 value might be used as complemen-tary exercise index in the suspicion and, perhaps, in the assessmentof the severity of pulmonary emphysema, in the light of a betterdefinition of the COPD phenotypes.

4.2. Ventilatory response to exercise

The secondary aim of the study was to characterize the V̇Eadaptation during maximal incremental CPET according with the

severity of emphysema quantified by lung HRCT. In patients withCOPD an increased ventilatory response is typically observed dur-ing both cycling and walking incremental exercise (Palange et al.,2000) and a steeper V̇E/V̇CO2 slope and a higher V̇E/V̇CO2 value at LT

172 P. Paoletti et al. / Respiratory Physiology &

Fig. 2. Correlation between %E quantified by HRCT and %EEST. %EEST was obtained bythe equation: %E = 58.1 + 11.9 × V /FEV − 0.8 × �V̇ /�V̇ . Dashed linesr

stwcca2

evarsm

Vcweraetpbea

odoamaPwt

rosi

to exercise in patient with emphysema quantified by HRCT. We

EST Tpeak 1 E CO2

epresent 95% confidence interval.

hould be expected. On the contrary, we found an inverse correla-ion between the %E and the V̇E/V̇CO2 slope. In our study, patientsith milder emphysema (i.e. %E ≤ 50) had a higher V̇E/V̇CO2 slope

ompared with control subjects (33 ± 5 vs 26 ± 3, p = 0.009) but,ontrarily to what expected, more severe patients (i.e. %E ≥ 50) hadV̇E/V̇CO2 slope similar to age-matched healthy controls (27 ± 3 vs6 ± 3, p = 0.3).

Why was the V̇E/V̇CO2 slope lower in patients with more severemphysema than in those with less severe emphysema (27 ± 3s 33 ± 5, p = 0.003)? Potential contributory factors include: (1)n improvement of ventilatory efficiency during exercise due toeduced physiological dead space; (2) a higher arterial CO2 (PaCO2 )et-point; (3) an abnormal dynamic respiratory mechanics whichay limit the physiological rise of V̇E in response to V̇CO2 .The detection of a higher PETCO2 peak in association with lower

˙ E/V̇CO2 slopes and ratios values at peak exercise make unlikely theontribution of a reduced physiological dead space (i.e., reducedasted ventilation) to the improved ventilatory efficiency during

xercise in our patients with more severe emphysema. There is noeason to believe that both the ability to increase lung perfusionnd to distribute inspired ventilation throughout the lungs duringxercise could be compromised (ventilation/perfusion inequalities)o a greater extent in patients with less severe emphysema com-ared with those with more severe emphysema. This contention isolstered by the finding that peak-to-rest change in VD/VT duringxercise did not significantly differ between patients with %E ≤ 50nd those with %E ≥ 50 (Table 4).

The state of “relative hypoventilation” (e.g., relative to the levelf CO2 produced) could in part be related to high PaCO2

valuesuring exercise. In this regard, the higher peak PETCO2 values inur patients with more severe emphysema (Table 3) might be due,t least in part, to a rise in airway resistances due to the well knownechanical derangements of the airways. Measurements of PaCO2

t peak exercise and, more importantly, peak-to-rest change inaCO2 during exercise clearly pointed out that patients with %E ≥ 50ere prone to increase PaCO2 during CPET more than their coun-

erpart with %E ≤ 50 (Table 4).In more severe emphysema (%E ≥ 50) the reduced exercise V̇E

elative to metabolic demand (V̇ ) may reflect reduced output

CO2f the central controller (“won’t breathe” hypothesis), or a pre-erved or amplified central respiratory drive in the presence of anmpaired mechanical/ventilatory muscle response (“can’t breathe”

Neurobiology 179 (2011) 167– 173

hypothesis) (Altose et al., 1977; Light et al., 1988; DeTroyer et al.,1993). Under these two circumstances CO2 retention would appear,consistently with the rise of PETCO2 and the greater peak-to-restchange in PaCO2 found in our patients with %E ≥ 50 during exercise(Tables 3 and 4). It has further been postulated that CO2 reten-tion during exercise may result from the “behavioural” adoptionof a shallow breathing pattern, which would serve to minimizeintrathoracic pressure perturbations, reduce respiratory discom-fort, and possibly obviate the development of respiratory musclefatigue (“wise fighter” hypothesis) (Begin and Grassino, 1991).Theoretically, the imbalance between inspiratory muscle load andcapacity may predispose patients with COPD to inspiratory musclefatigue or frank task failure and consequent CO2 retention (Beginand Grassino, 1991). However, it has recently been shown thatmaximal inspiratory pressure generation (Montes de Oca and Celli,2000), the pattern of ventilatory muscle recruitment (Montes deOca and Celli, 2000), and breathing pattern (Montes de Oca andCelli, 2000; O’Donnell et al., 2002) were not different in those whomaintained eucapnia and those who developed hypercapnia duringexercise, thereby casting doubt on fatigue, or its avoidance by theadoption of a rapid shallow breathing pattern, as the explanation forCO2 retention. Accordingly, no significant differences in breathingpattern were found between patients with more severe emphy-sema (%E ≥ 50) and those with less severe emphysema (%E ≤ 50) inour study.

Another possible explanation is the developing of lung mechan-ical constraint during exercise (i.e., dynamic lung hyperinflation)that may limit the physiological rise of V̇E in more severe emphy-sema patients (%E ≥ 50), therefore contributing to a higher peakPETCO2 values and a greater peak-to-rest change in PaCO2 duringexercise (Tables 3 and 4). In this regard during incremental CPET thepresence of expiratory flow-limitation may lead to an inappropriateincrease of VT so that patient increases the V̇E mostly by increasingfR (shallow and rapid breathing). However, we did not detect anydifference in VT expansion (in absolute value and as % of either FVCor SVC) or fR between patients with %E ≥ 50 and those with %E ≤ 50;in addition no correlation was found between fR and %E (Table 3).Also, the presence of expiratory flow-limitation would promote theincomplete lung emptying which, in turn, would force patients tobreathe at progressively increased lung volumes (i.e., dynamic lunghyperinflation) with resultant increased oxygen cost of breath-ing, thus imposing an early mechanical limitation/constraint to V̇E(O’Donnell et al., 2001; Laveneziana et al., 2011). In this regard,although dynamic lung hyperinflation was present in the majorityof our patients, the extent of it was significantly less in patientswith %E ≤ 50% than in those with %E ≥ 50% (−0.25 ± 0.31 l vs−0.94 ± 0.38 l, p = 0.01) (Table 4). In addition, indices of progressivemechanical/volume constraints such as VT/IC ratio at peak exerciseand peak-to-rest change in VT/IC ratio (�VT%IC) during exercise,were significantly greater in more severe emphysema group thanin less severe emphysema group (96 ± 5 vs 81 ± 6, p = 0.003 and72 ± 6 vs 49 ± 10, p = 0.002, respectively) (Table 4). Thus, it is plausi-ble that greater rates of dynamic lung hyperinflation in associationwith greater attainment of critical volume constraints on exer-tion (VT/IC ratio at peak and �VT%IC, Table 4) may have limitedthe ability to increase V̇E in response to the increasing metabolicdemands of exercise in the more severe emphysema patients(i.e., %E ≥ 50%).

4.3. Conclusions

This study is the first to characterize the ventilatory response

were able to demonstrate that a distinct physiologic response pat-tern can be detected at CPET in these patients. The ratio VT peak/FEV1progressively increased in proportion to the severity of emphysema

logy &

a%vsm

C

A

L

R

A

B

B

B

C

D

G

H

K

L

P. Paoletti et al. / Respiratory Physio

nd a ratio > 1 was typically observed in more severe patients (i.e.,E ≥ 50 at HRCT). In addition, the V̇E/V̇CO2 slope and the PETCO2 peakalues decreased and increased respectively as more as the emphy-ema was severe. Finally, the VTpeak

/FEV1 values and V̇E/V̇CO2 slopeight be used to predict the severity of emphysema.

onflict of interest statement

The authors declare that they have no conflict of interest.

cknowledgment

This study was supported by a grant of “Fondazione Eleonoraorillard Spencer Cenci, Rome, Italy”.

eferences

ltose, M.D., McCauley, W.C., Kelsen, S.G., Cherniack, N.S., 1977. Effects of hyper-capnia and inspiratory flow-resistive loading on respiratory activity in chronicairways obstruction. J. Clin. Invest. 59, 500–507.

eaver, W.L., Wasserman, K., Whipp, B.J., 1986. A new method for detecting theanaerobic threshold by gas exchange. J. Appl. Physiol. 60, 2020–2027.

egin, P., Grassino, A., 1991. Inspiratory muscle dysfunction and chronic hypercapniain chronic obstructive pulmonary disease. Am. Rev. Respir. Dis. 143, 905–912.

lackie, S.P., Fairbarn, M.S., McElvaney, N.G., Wilcox, P.G., Morrison, N.J., Pardy, R.L.,1991. Normal values and ranges for ventilation and breathing pattern at maximalexercise. Chest 100, 136–142.

ampbell, S.C., 1982. A comparison of the maximum volume ventilation with forcedexpiratory volume in one second: an assessment of subject cooperation. J. Occup.Med. 24, 531–533.

eTroyer, A., Leeper, J.B., McKenzie, D., Gandevia, S., 1993. Neural drive to thediaphragm in patients with severe COPD. Am. J. Respir. Crit. Care Med. 155,1335–1340.

ietema, H.A., Zanen, P., Schilham, A., van Ginneken, B., van Klaveren, R.J., Prokop,M., Lammers, J.W., 2010. Distribution of emphysema in heavy smokers: impacton pulmonary function. Respir. Med. 104, 76–82.

offman, E.A., Jiang, R., Baumhauer, H., Brooks, M.A., Carr, J.J., Detrano, R., Reinhardt,J., Rodriguez, J., Stukovsky, K., Wong, N.D., Barr, R.G., 2009. Reproducibility andvalidity of lung density measures from cardiac CT scans—The Multi-Ethnic Studyof Atherosclerosis (MESA) Lung Study. Acad. Radiol. 16, 689–699.

urashima, K., Takayanagi, N., Sato, N., Kanauchi, T., Hoshi, T., Tokunaga, D., Ubukata,

M., Yanagisawa, T., Sugita, Y., Kanazawa, M., 2005. High resolution CT andbronchial reversibility test for diagnosing COPD. Respirology 10, 316–322.

aveneziana, P., Valli, G., Onorati, P., Paoletti, P., Ferrazza, A.M., Palange, P., 2011.Effect of heliox on heart rate kinetics and dynamic hyperinflation during high-intensity exercise in COPD. Eur. J. Appl. Physiol. 111, 225–234.

Neurobiology 179 (2011) 167– 173 173

Light, R.W., Mahutte, C.K., Brown, S.E., 1988. Etiology of carbon dioxide retention atrest and during exercise in chronic airflow obstruction. Chest 94, 61–67.

Makita, H., Nasuhara, Y., Nagai, K., Ito, Y., Hasegawa, M., Betsuyaku, T., Onodera, Y.,Hizawa, N., Nishimura, M., Hokkaido COPD Cohort Study Group, 2007. Charac-terisation of phenotypes based on severity of emphysema in chronic obstructivepulmonary disease. Thorax 62, 932–937.

Miller, M.R., Hankinson, J., Brusasco, V., Burgos, F., Casaburi, R., Coates, A., Crapo, R.,Enright, P., van der Grinten, C.P., Gustafsson, P., Jensen, R., Johnson, D.C., MacIn-tyre, N., McKay, R., Navajas, D., Pedersen, O.F., Pellegrino, R., Viegi, G., Wanger,J., 2005. ATS/ERS task force standardisation of spirometry. Eur. Respir. J. 26,319–338.

Montes de Oca, M., Celli, B., 2000. Respiratory muscle recruitment and exercise per-formance in eucapnic and hypercapnic severe chronic obstructive pulmonarydisease. Am. J. Respir. Crit. Care Med. 61, 880–885.

Omori, H., Fujimoto, K., Katoh, T., 2008. Computed-tomography findings of emphy-sema: correlation with spirometric values. Curr. Opin. Pulm. Med. 14, 110–114.

O’Donnell, D.E., Revill, S.M., Webb, K.A., 2001. Dynamic hyperinflation and exerciseintolerance in chronic obstructive pulmonary disease. Am. J. Respir. Crit. CareMed. 164, 770–777.

O’Donnell, D.E., D’Arsigny, C., Fitzpatrick, M., Webb, K.A., 2002. Exercise hypercapniain advanced COPD: the role of lung hyperinflation. Am. J. Respir. Crit. Care Med.166, 663–668.

Palange, P., Forte, S., Onorati, P., Manfredi, F., Serra, P., Carlone, S., 2000. Ventilatoryand metabolic adaptations to walking and cycling in patients with COPD. J. Appl.Physiol. 88, 1715–1720.

Palange, P., Valli, G., Onorati, P., Antonucci, R., Paoletti, P., Rosato, A., Manfredi, F.,Serra, P., 2004. Effect of heliox on lung dynamic hyperinflation, dyspnea, andexercise endurance capacity in COPD patients. J. Appl. Physiol. 97, 1637–1642.

Palange, P., Ward, S.A., Carlsen, K.H., Casaburi, R., Gallagher, C.G., Gosselink, R.,O’Donnell, D.E., Puente-Maestu, L., Schols, A.M., Singh, S., Whipp, B.J., ERS TaskForce, 2007. Recommendations on the use of exercise testing in clinical practice.Eur. Respir. J. 29, 185–209.

Park, K.J., Bergin, C.J., Clausen, J.L., 1999. Quantification of emphysema with three-dimensional CT densitometry: comparison with two-dimensional analysis,visual emphysema scores, and pulmonary function test results. Radiology 211,541–547.

Quanjer, P.H., Tammeling, G.J., Cotes, J.E., Pedersen, O.F., Peslin, R., Yernault, J.C., 1993.Lung volumes and forced ventilatory flows. Report Working Party Standardiza-tion of Lung Function Tests, European Community for Steel and Coal OfficialStatement of the European Respiratory Society. Eur. Respir. J. Suppl. 16, 5–40.

Vogiatzis, I., Georgiadou, O., Golemati, S., Aliverti, A., Kosmas, E., Kastanakis, E.,Geladas, N., Koutsoukou, A., Nanas, S., Zakynthinos, S., Roussos, C., 2005. Pat-terns of dynamic hyperinflation during exercise and recovery in patients withsevere chronic obstructive pulmonary disease. Thorax 60, 723–729.

Wanger, J., Clausen, J.L., Coates, A., Pedersen, O.F., Brusasco, V., Burgos, F., Casaburi, R.,Crapo, R., Enright, P., van der Grinten, C.P., Gustafsson, P., Hankinson, J., Jensen,

R., Johnson, D., Macintyre, N., McKay, R., Miller, M.R., Navajas, D., Pellegrino, R.,Viegi, G., 2005. Standardisation of the measurement of lung volumes. Eur. Respir.J. 26, 511–522.

Wasserman, K., Hansen, J.E., Sue, D.Y., 1994. Principles of Exercise Testing and Inter-pretation. Lea & Febiger, Philadelphia.