Measurement of ‘closing volume‘ initiated from functional residual capacity

Respiration Physiology (1976)28, 325-33 1 0 Elsevier/North-Holland Biomedical Press

MEASUREMENT OF ‘CLOSING VOLUME’ INITIATED FROM FUNCTIONAL RESIDUAL CAPACITY’

D. B. CRAIG, D. S. MCCARTHY, I. J. GILMOUR, R. M. CHERNIACK Departments

of Medicine

and Anaesthesia,

Investigation

University

of Manitoba,

Unit, Health Sciences Centre,

L. D. H. WOOD and

and the Respiratory

Winnipeg, Manitoba,

Division,

Clinical

Canada

Abstract. Comparison of the nitrogen method closing volume (CV) test, with oxygen inspiration initiated at residual volume (RV method) and functional residual capacity (FRC method), was made in YI seated normal subjects. For RV and FRC methods, respectively CV%VC (mean + SD) was 14.4”,, (k6.2) and 17.5:‘; (k7.5) (P = 0.005); slope of Phase III of CV trace was 0.99% NJ1 (kO.76) and 1.66”;, NJ1 (+ I .07) (P = 0.005); size of cardiogenic oscillations was 1.057, N, (kO.42) and I.21 ‘?J”N, (kO.40) (P = 0.001). These data confirm earlier predictions, based on a calculated increased lung top to bottom N, gradient in the FRC method. Support for this mechanism was obtained in 5 additional normal subjects in whom the increased CV:,VC, slope of Phase III and size of cardiogenic oscillations with the FRC method were eliminated when the top-to-bottom N, gradient was reduced by breathing a reduced FI,~. Measurements made using the classical RV method cannot be directly compared to those using the FRC method. Airway closure Cardiogenic oscillations

Closing volume Sequential lung emptying

The single breath nitrogen closing volume (CV) test, as first described by Anthonisen et al. (1969) requires vital capacity (VC) inspiration of oxygen from residual volume

(RV), followed by slow expiration back to RV. Because of differences in regional RV and VC, this maneuver results in a nitrogen gradient in the lung at total lung capacity (TLC) with maximum N, concentrations in upper regions and minimum in lower regions. Although subsequent widespread use of the N, CV test has retained the VC inspiration of O2 from RV, (RV method) an inspiratory capacity breath of

Acceptedfor

publication 2 September

1976.

’ From the Collaborative Study of Smoking and Airways Obstruction, Supported by a Contract (# NHLI 73-2902R), National Heart & Lung Institute, Bethesda, Maryland. Supported in part by the Medical Research Council of Canada. 325

326

D. B. CRAIG

et al.

0, from FRC (FRC method) will also produce a top to bottom N, gradient, due to differences in regional FRC and IC. Kaneko et al. (1975), using a lung model, predicted that the CV’AVC measured by the FRC method should be larger than that using the RV method. The basis for this hypothesis was the calculated increase in the magnitude of the top to bottom N, gradient with the FRC start, which would make the terminal slope change in the CV curve more discrete and therefore detectable at an earlier point, leading to the measurement of an increased CVXVC. The same authors confirmed their lung model predictions by demonstrating an increased CV%VC when an air-containing dead space was added to the inspiratory breathing circuit during CV measurement by the RV method. Although this had the effect of causing 0, inspiration to begin at a lung volume near FRC, they did not directly compare RV and FRC methods. In contrast, Hyatt et al. (1973) in 3 seated subjects found no difference in CV%VC using the FRC method, as compared to the RV method. Hedenstierna et al. (1976) compared RV and FRC methods in 5 supine subjects, and while not commenting on actual values of CV”/OVC, noted with the FRC method that the transition between Phases III and IV of the CV trace was less well defined, in disagreement with the predictions of Kaneko et al. (1975). Since there has been no direct comparison of RV and FRC methods except in a limited number of subjects, and in view of the disagreements in the literature outlined above, we administered the N, CV test, using RV and FRC methods, to 91 normal subjects. In addition to comparing CV%VC with the two methods, we have also examined the slope of Phase III of the CV trace, and the size of the cardiogenic oscillations. These latter comparisons have not previously been reported. In a separate group of normal subjects we have obtained further support for Kaneko’s hypothesis that the origin of the difference in CV%VC between the RV and FRC methods is a change in the magnitude of the top to bottom N, concentration gradient.

Methods

Ninety-one hospital employees who denied symptoms of respiratory disease volunteered to take part in the study. There were 33 males, 58 females, 44 smokers, 47 nonsmokers. Mean age was 36.5 years (SD 12.4); 55 subjects were less than 40 years old, 36 were over 40 years. All studies were performed with the subjects seated and, except for initiation of oxygen inspiration from FRC, the procedures used to obtain and analyze CV traces were according to the criteria suggested by the National Institute of Health (1973). The onset of Phase IV was taken as the point of final departure from the best-fit line drawn by eye through the latter half of Phase III. The order of CV measurement with inspiration from FRC or RV was varied in a random fashion. The mean of two or three satisfactory traces for each of the FRC and RV method was taken. Examples of tracings for one subject using the two methods are seen in

‘CLOSING

VOLUME’

INITIATED

FROM FUNCTIONAL

terminal

by RV method slope change

(bottom) in alveolar

CAPACITY

327

litres

VOLUME

Fig. I. CV tracings

RESIDUAL

and FRC method plateau.

(top) in same subject.

CV = volume

Arrow

indicates

from this point to RV.

fig. 1. CV, i.e. the lung volume between the point of inflection of the N, trace and RV, was calculated and expressed in liters and as a percent of VC. The slope of Phase III of the nitrogen trace was estimated from the line of best fit along the alveolar plateau (ignoring the first 30 7; of VC expired) as the % N, per liter expired. The mean height of the five largest cardiogenic oscillations after 30 % of the vital capacity had been expired was determined. Expired VC was measured directly. Data derived from the two techniques were compared by r-test analysis for paired data. A separate group of five normal seated subjects breathed a reduced inspired nitrogen concentration (FI~J through a nonrebreathing valve until the FEN* was stable. The FI,, was adjusted for each subject (mean FIN* = 0.45) so that the mean FEN, using the FRC method was similar to the mean FEDSusing the RV method, during air breathing. Decreasing the FIEF and therefore the FADEwould decrease the top to bottom N, gradient (AN,) at TLC following 0, inspiration during the CV maneuvers. The estimated FADEin areas 10 and 25 cm from lung apex was calculated for each of the three maneuvers (fig. 2). The mean data of Milic-Emili et al.

Fig. 2. Diagram upright

TLC considered methods,

representing

lung units at top (10 cm from apex) and bottom

lung. Mean data of Milk-Emili equal.

Estimated

FAN, and AN,

with FINl = 0.79 and FRC method

(FI,,

et al. (1966) was used for regional method

with

= 0.79) and decreases

in the two lung regions Fr,,

= 0.45. Estimated

with FRC method

(25 cm from apex) of an

RV and FRC, is shown AN,

at a reduced

with regional

for RV and FRC

increases R,,.

with FRC

328

D. B. CRAIG

et a/.

(1966) was used for regional RV and FRC. Regional TLC’s were assumed equal, and, for convenience, FADEwas assumed equal to FIEF. Estimated regional FNr after 0, inspiration to TLC was obtained from the formula:

FrN2(TLC) = (Vr/TLCr) FrN2 Vr and FrNz indicate initial regional lung volume and N, concentration, and TLCr and FrN2 (TLC) regional lung volume and N, concentration at TLC. When 0, inspiration during the CV maneuver began at RV, the estimated top-to-bottom AN, at TLC was 10.4 ‘?/,,with 0, inspiration from FRC, the estimated AN, was 14.0 9,; (FI,~ = 0.79) and 8.1 7; (FlN2 = 0.45).

Results

Table 1 compares results obtained from the two maneuvers in the 9 1 subjects. When oxygen inspiration began at FRC, CV (1) CV%VC, the slope of Phase III, and the size of the cardiogenic oscillations were all significantly larger. Expired VC was similar in the two methods. The significant differences between the RV and FRC maneuvers remained when analysis was repreated with subjects separated as to age (< 40 and > 40) sex, or smoking history. TABLE Results

1

of N, CV test using RV and FRC methods RV 4.40

vc (1) cv”,,vc

14.4

CV (1) Phase III slope (y,, NJl) Cardiogenic Mean

oscillations

(N,“,)

(I .OO) (6.2)

4.44 (1.03) 17.5

NS

(7.5)

0.005

0.62 (0.25)

0.76 (0.32)

0.99 (0.76)

1.66 (1.07)

0.005 0.005

1.05 (0.42)

1.21 (0.40)

0.001

+ SD, 91 subjects.

* from paired

t-test. TABLE

Comparison

2

of effect of initiation of 0, inspiration from RV (A) and FRC (B) following from FRC (C) while breathing 45”1,, N2* in 0, Panel in

F’h2

CV” 0 vc

fig. 2

Mean

P*

FRC

Slope

Cardiogenic

Phase III

oscillations

(“A N#)

(‘i, N,) 0.93 10.63 1.31,0.41

RV

(A)

0.79

14.2 +_1.9

1.10+0.91

FRC FRC

(B) (C)

0.79 0.45*

16.2k2.4

I .71 * 1.03 0.85 +0.47

k SD, 5 subjects.

* Mean FI,~.

14.0*1.9

air breathing

0.66 +0.29

and

‘CLOSING VOLUME’ INITlATED FROM FUNCTIONAL RESIDUAL CAPACITY

329

Table 2 summarizes the results in the live additional subjects in whom the FI,~ was reduced prior to one series of FRC maneuvers. As in the larger group, CV%VC, the slope of Phase III, and the size of the cardiogenic oscillations were larger when 0, inspiration began at FRC, during air breathing. When the FIEF was reduced to a mean of 0.45, thereby reducing the estimated top to bottom AN,, the increases which had been present, disappeared.

The observed increase in CV”/,VC using the FRC method confirms the prediction of Kaneko et al. (1975). As they had also noted following addition of an air-containing dead space, and using the RV method, we found that tracings generated by the FRC method presented less difficulty in analysis, because the slope change between Phase III and IV was more distinct. Further support for the hypothesis that an increased top to bottom AN, in the FRC method causes the increase in CV%VC is provided by our data from the FRC method during breathing a reduced FIN>. Reduction in estimated AN, from 14 to 8 % was accompanied by the return of the increased CVo!VC to the lower value obtained using the RV method. As suggested by Kaneko et al. (1975) the reason for the association of an increased CV% VC with an increased d N, is most likely the result of an improved ability to detect the slope change between Phases III and IV. The phenomenon producing this slope change is not itself changing in these circumstances, but the ability to detect it by the CV test does change. The observations of Hyatt et al. (1973) and Hedenstierna et al. (1976) differ from the findings of the present study, as outlined earlier. These differences are likely a reflection of the small number of subjects studied by both previous groups. Although Hedenstierna’s subjects were studied while supine, this is unlikely to be the cause of the reported differences, since regional lung volume differences exist due to gravitational influences, regardless of body position. In addition, studies in supine subjects in our laboratory have found CV”/,VC greater using the FRC method. The increases in the slope of Phase III of the CV trace and the size of the cardiogenic oscillations with the FRC method have not previously been reported, The size of cardiogenic oscillations observed in the CV trace are thought to be mainly the result of top-to-bottom differences in regional nitrogen concentration, and the cardiac stroke volume (Engel et al., 1973). Since the FRC maneuver increases the topto-bottom AN,, the observed increase in size of cardiogenic oscillations is expected. The reduction in the size of cardiogenic oscillations during the FRC maneuver when the AN, was reduced by the decreased FI,, is further support for this explanation. The positive slope of Phase III of the CV trace had been thought to result from regional and stratified inhomogeneity (Buist and Ross, 1973). Inhomogeneity between lung regions results from compliance-related differences in regional volume changes during the breathing maneuvers of the CV test. Stratified inhomogeneity

330

D. B. CRAIG et al.

is thought to be due to incomplete diffusion and convective mixing in peripheral lung units. More recently, however, Engel et al. (1974) recorded qualitatively similar alveolar plateaux in dogs’ tracheas and peripheral airways. They demonstrated that in addition to variations in ventilation per unit volume (dV/VO) between lung regions, variations within regions contribute to the positive slope of the alveolar plateau seen following a single breath of oxygen. Our observations of a direct correlation between the slope of Phase III and the estimated interregional AN, is suggestive but not conclusive evidence that this slope is in fact dependent upon an apex to base N, gradient. Within lung regions, stratified inhomogeneity would be expected to be increased in the FRC method, due to the smaller volume change (IC rather than VC) which would also occur over a shorter time period, allowing less time for gas mixing. While we suggest that the observed differences between the results of the RV and the FRC method CV tests were due mainly to the changes in interregional AN,, intraregional variations in AN, are also likely. If intraregional variations in RV/TLC and FRC/TLC ratios are similar to interregional variations, intraregional differences in AN, in RV and FRC methods can be expected. The observed increases in CV% VC, slope of Phase III and the size of cardiogenic oscillations clearly indicate that the FRC modification of the CV test must be considered a different test than the classical method. Data obtained using the FRC version therefore cannot be compared directly to the large body of existing data obtained from the RV version. Since the FRC version is a different test, any attempt to apply it to the general population would have to be preceded by a careful reexamination of all of the factors that have been found to influence the RV version of the test. In addition, new normal standards would have to be developed. We recognize no indication to proceed in this fashion. The FRC modification does have some valid applications however. It makes possible the measurement of CV by the N, method in anesthetized patients, and in patients with restricted vital capacities due to pain. These data can, of course, only be compared with data obtained in the same fashion. We conclude that the FRC method N, CV test produces results different than the classical RV method. CV%VC, the slope of Phase III and the size of the cardiogenie oscillations are all increased. These differences appear due mainly to an increase in interregional top-to-bottom N, concentration difference. Measurements made using the RV method cannot be directly compared to those using the FRC method.

Acknowledgements

The expert assistance of Mr. Peter West with the collection and analysis of the data is gratefully acknowledged.

‘CLOSING

VOLUME’ INITIATED

FROM FUNCTIONAL

RESIDUAL

331

CAPACITY

References Anthonisen,

N. R., J. Danson,

of age. Respir. Physiol.

P. C. Robertson

and W. R. D. Ross (1969). Airway

Buist, A. S. and B. B. Ross (1973).,Quantitative airway

obstruction. holding

analysis

L. D. Wood,

studied

G. Utx, J. Joubert

by intrapulmonary

lung units. J. Appl. Physiol. G., G. McCarthy

Anesthesiology Hyatt,

R. E., G. C. Okeson

Milic-Emili, National

and M. Bergstrom

and J. R. Rodarte

and 0. Balchum

by nitrogen

method.

of early

(1973). Gas mixing during

J. Appl. Physiol.

(1974). Ventilation

(1973). Influence

35 : Y-1 7. distribution

closure during

in anatomical

mechanical

ventilation.

Volume

M. B. Dolovich, Bethesda,

Determinations

ofexpiratory

on the pattern

lung volume

on closing

volume

38: 10-15.

D. Tropand

K.Kaneko

(lY66). Regional

distribution

21 : 749-759.

Maryland

(Nitrogen

flow limitation

35: 411419.

(1975). Effect of preinspiratory J. Appi. Physiol.

gas in the lung. J. Appl. Physiol.

Heart and Lung Institute,

for Closing

and P. T. Macklem

(1976). Airway

man. J. Appl. Physiol.

J., J. A. M. Henderson,

of inspired

in the diagnosis

37: 194200.

in normal

K., J. Mohler

determination

plateau

44: 114123.

of lung emptying Kaneko,

of the alveolar

gas sampling.

Engel, L. A., G. Utx, L. D. H. Wood and P. T. Macklem Hedenstierna,

as a function

Am. Rec. Respir. Dis. 108: 1078-1087.

Engel, L. A., H. Menkes, breath

closure

8: 58-65.

(July IY73), Suggested

Method).

Standardized

Procedures