We investigated the differential effect of histamine and methacholine on spirometry and ventilation distribution (where indexes S cond andS acin represent conductive and acinar ventilation heterogeneity; Verbanck S, Schuermans D, Van Muylem A, Noppen M, Paiva M, and Vincken W. J Appl Physiol 83: 1807–1816, 1997). Thirty normal subjects were challenged with cumulative doses of 6.52 μmol histamine and, on a separate day, with either 6.67 μmol methacholine (equal-dose group; n = 15) or 13.3 μmol methacholine (double-dose group; n = 15). Largest average forced expiratory volume in 1 s (FEV1) decreases or S cond increases obtained in either group were −9% and +286%, respectively;S acin remained unaffected at all times. In the equal-dose group, a smaller FEV1 decline (P= 0.002) after methacholine was paralleled by a smallerS cond increase (P = 0.041) than with histamine. However, in the double-dose group, methacholine maintained a smaller FEV1 decline (P = 0.009) while inducing a larger S cond increase (P = 0.006) than did histamine. The differential action of histamine and methacholine is confined to the conductive airways, where histamine likely causes the greatest overall airway narrowing and methacholine induces the largest parallel heterogeneity in airway narrowing, probably at the level of the large and small conductive airways, respectively. The observed ventilation heterogeneities predict a risk for dissociation between ventilation-perfusion mismatch and spirometry, particularly after methacholine challenge.
- airway narrowing
- nitrogen washout
- parallel heterogeneity
bronchial hyperresponsiveness is usually evaluated by provocation testing with histamine or methacholine, in connection with a measurement of forced expiratory volume in 1 s (FEV1). Although guidelines for provocation testing suggest that methacholine and histamine can be used interchangeably for routine clinical practice (21), different mechanisms are thought to underlie the airway narrowing process induced by these agents. Methacholine is thought to act directly on muscarinic receptors on bronchial smooth muscle. Histamine may act through a direct effect on bronchial smooth muscle (6), probably via stimulation of H1receptors (5), or indirectly as an irritant stimulating vagal parasympathetic reflex bronchoconstriction (20), although the former effect is thought to dominate over the latter (14). In any case, different mechanisms appear to be reflected in the differential behavior of histamine and methacholine, at least in normal human subjects (22, 26), in whom histamine elicits the greatest response in terms of FEV1. In asthmatic subjects, the differential behavior of histamine and methacholine is less clear: when considering all dosages in moles (27), some studies do show a difference (7,11) whereas others do not (2, 10, 18, 26). These conflicting reports in asthmatic subjects may be due to the degree to which some of the protective mechanisms against provocation are deficient in any particular group of asthma patients, e.g., the protective effect of the inhibitory muscarinic receptors present in normal subjects may dysfunction to a greater or lesser degree in asthmatic subjects (1). Such findings underline the need to fully understand the differential action of histamine and methacholine in the normal lung.
It has been argued that the inability to reveal differences between histamine and methacholine in some studies may have been related to the intrinsic noise of the measurement (2, 7) and the gross averaging of lung function indexes such as FEV1(3). In contrast to spirometry, which is generally considered to reflect a modification of airway caliber and elastic properties of the lungs as a whole, ventilation distribution tests rely on parallel differences in airway caliber and elastic properties of subtended lung units. We have previously shown how indexes of ventilation heterogeneity originating in the conductive airway zone (S cond) and acinar airway zone (S acin), as derived from a multiple-breath washout (MBW) test, are affected by histamine provocation (25). On the basis of the very markedS cond increases and the absence ofS acin increases observed in hyperresponsive and in nonhyperresponsive subjects (classified according to the 20% FEV1 criterion), we concluded that histamine induced 1) differences in specific ventilation of lung units larger than acini and 2) flow asynchrony between these units during exhalation. In the histamine provocation study (25), S cond more or less paralleled the FEV1 behavior across both groups, indicating that there was a general decrease of airway lumen as well as an inequality in narrowing of parallel airways of the same generation, probably caused by inequalities in bronchial tone of the conductive airways. In the present study, we compared methacholine with histamine in terms of respective spirometry and ventilation heterogeneity in normal subjects.
MATERIALS AND METHODS
Two groups of 15 subjects underwent methacholine provocation testing after having been classified as nonhyperresponsive on the basis of a decrease in FEV1 <20% baseline after a cumulative dose of 2 mg histamine (i.e., 6.52 μmol). All subjects were recruited on a voluntary basis in the framework of a prevalence study on hyperresponsiveness in normal subjects in the Brussels area (using a maximum cumulative dose of 2 mg histamine by default). The methacholine procedure always followed the histamine procedure, separated by at least 1 wk and by no longer than 1 mo. The first group (equal-dose group) received the same cumulative dose of 6.67 μmol methacholine (i.e., 1.6 mg) whereas the second group (double-dose group) included one more doubling dose only in the methacholine provocation protocol, to yield a cumulative dose of 13.33 μmol (i.e., 3.2 mg). By using the dosimeter technique (MEFAR dosimeter MB3; vital capacity breath), histamine was administered in four steps (0.52, 1.56, 3.39, and 6.52 μmol) and methacholine in four or five steps (0.83, 1.67, 3.34, 6.67, and 13.3 μmol). On both occasions, subjects were free of respiratory symptoms.
Spirometric parameters were obtained by means of standardized lung function laboratory equipment (Sensormedics, Bilthoven, Netherlands) according to recommended procedures. MBW tests were carried out with a dedicated breathing assembly identical to the one described in detail in our previous work (25). In brief, this equipment consists of a pneumatic valve system that allows a sequence of 20–25 1-liter inspirations from a bag filled with pure O2 and exhalations through a separate pathway. Throughout the test, end-inspiratory volume was functional residual capacity (FRC), and the exact number of breaths was dependent on subject performance and on progressive N2 dilution. Volume and N2 concentration were acquired at 25 Hz. One MBW run typically takes 2–3 min and is separated by 3–5 min before a subsequent MBW maneuver is started, to allow the subject to wash out completely the inhaled O2.
Before starting the provocation testing, subjects performed spirometry, involving three forced expiratory maneuvers, from which the best maneuver was selected to provide baseline FEV1, peak expiratory flow (PEF), and forced expiratory flow after exhaling 75% of the forced vital capacity (FEF75), and three baseline MBW tests. At each successive step of the provocation procedure with histamine (4 steps) or methacholine (4 or 5 steps), spirometry was repeated. For operational reasons, MBW tests could not be carried out at each intermediate step of the provocation procedures. In the equal-dose group, subjects performed two MBW tests immediately after spirometry after the final dose of either histamine (6.65 μmol) or methacholine (6.67 μmol). In the double-dose group, subjects performed two MBW tests immediately after spirometry after the final dose of either histamine (6.65 μmol) or methacholine (13.3 μmol). To account for possible time-dependent effects occurring over the course of the two final-dose MBW tests, these were followed by another spirometry. In this way, spirometric data obtained for the final dose of histamine or methacholine were the average values of FEV1, PEF, and FEF75 obtained immediately before and after the two final-dose MBW tests.
Because of the importance in understanding the computation of some MBW-derived indexes and how they relate to the mechanisms of ventilation heterogeneity, we reiterate some basic features of the MBW phase III slope analysis. In our previous work (25), a more extensive description and theoretical background of all MBW-derived indexes can be found. Briefly, N2 phase III slope is computed by linear regression between 0.65 liter and end of expiration (nominally 1 liter) and divided by the mean expired N2 concentration of that expiration in each subsequent breath of the MBW test. In this way, the normalized N2phase III slope represents relative N2 concentration differences in different parts of the lungs with respect to the prevailing N2 concentration in the lungs at any instance of the MBW test. Normalized phase III slopes for each breath are then plotted against lung turnover (TO), i.e., cumulative expired volume corresponding to each subsequent breath divided by the subject's FRC. Lung TO is used instead of breath number to account for possible differences in lung dilution due to differences in the subjects' FRC (8).
Figure 1 shows some typical normalized phase III slope plots obtained from the average of three baseline MBW tests (triangles) and the average of two provocation MBW tests (circles) obtained in one subject from the equal-dose group (Fig.1 A) and in another subject from the double-dose group (Fig.1 B). In both panels of Fig. 1, open symbols represent histamine data sets (baseline or provocation) and closed symbols represent methacholine data sets (baseline and provocation). From the average normalized slope curves, such as in Fig. 1,S cond and S acincontribution are derived as follows. The MBW indexS cond is computed as the normalized phase III slope difference per unit TO, determined in the part of the MBW in which only conductive airways are known to contribute to the rate of rise of the normalized phase III slope, e.g., between TO = 1.5 and TO = 6 (see Theoretical background). In the case of the subject in Fig. 1 A, baselineS cond = 0.032 liter−1 (for both histamine and methacholine protocols) whereas provocationS cond = 0.073 liter−1(histamine) and 0.051 liter−1 (methacholine). In the case of the subject in Fig. 1 B, baselineS cond = 0.024 liter−1 (for both histamine and methacholine protocols), whereas provocationS cond = 0.048 liter−1(histamine) and 0.076 liter−1 (methacholine).S acin is determined by subtracting from the slope of the first breath the part that can be attributed to the conductive airways, i.e., S cond in proportion to the TO value of the first breath. For the subjects of Fig. 1, respective S acin values were 0.07 and 0.05 liter−1 at all instances of the provocation procedures.
In addition to the normalized N2 phase III slope analysis, the decreasing mean expired N2 concentration in subsequent breaths of the MBW test (usually referred to as the N2-washout curve) is used for the computation of Curv, an index of the ventilation inhomogeneities, more precisely differences in specific ventilation, between different parts of the lung (seeTheoretical background). It can be computed as follows. From the semilog plot of mean expired N2concentration as a function of TO, Curv is determined as the ratio of the regression slope between TO = 3 and TO = 6 to the regression slope between TO = 0 and TO = 3. In this way, Curv is always smaller than or equal to 1, and a more curvilinear N2 washout curve leads to a smaller value for Curv. Finally, in the first expiration of each MBW test, anatomical dead space (VDanat) was determined according to Fowler's method.
S cond and S acinindependently reflect ventilation heterogeneities occurring at branch points located in the conductive and acinar lung zones, respectively (24, 25). In the present context of bronchoprovocation in normal subjects in whom conductive airways are expected to be the main component of ventilation heterogeneity (25), we focus on the principles underlying S cond. The gas transport mechanism leading to nonzero S condvalues even in normal lungs is convective gas flow. In fact,S cond is the most straightforward MBW index in that it directly relates to N2 concentration differences accumulating between any two lung units of which one lung unit is preferentially ventilated at every subsequent inhalation. The only condition for this concentration difference to translate into a sloping phase III during each subsequent exhalation is that there is a sequential emptying pattern between these lung units with the most ventilated one emptying preferentially at the onset of exhalation. In every subsequent breath, these concentration differences will accumulate with respect to the prevailing average concentration in the lung, and if the phase III slope is normalized by the mean expired concentration in each expiration, such a mechanism will cause the normalized slope to rise as in Fig. 1. If this were the only mechanism responsible for the normalized slopes, their extrapolation of the normalized slope curves to TO = 0 would be zero. This is obviously not the case (Fig. 1), and the initial slopes are in fact mainly due to the much more complex mechanism involving simultaneous convective and diffusive gas transport in the lung periphery (which underliesS acin). However, after only a few breaths (from TO = 1.5 onward) an equilibrium state of convection and diffusion is reached in which relative concentration differences remain constant (and produce a nearly constant contribution to the normalized phase III slope) throughout most of the MBW (16). That is why we compute S cond only in that part of the normalized alveolar slope curve in which only the conductive airway heterogeneity contributes to the rate of slope increase, i.e., beyond TO = 1.5.
In contrast to S cond, which requires sequential emptying to reflect differences in concentration between any two lung units, the curvilinearity of the N2 mean expired concentration washout curve represents concentration (specific ventilation) differences between lung units even if both units empty synchronously. In other words, S cond is different from 0 only when lung units with different inspired gas concentrations empty asynchronously, whereas Curv becomes smaller than 1 in the case of gas concentration differences between lung units, virtually irrespective of emptying patterns. The smaller Curv reflects increased heterogeneity of specific ventilation between any two lung units. Although both S cond and Curv are thought to originate at branch points in the conductive airway tree, where convective transport prevails, it is a priori not possible to distinguish between levels within the conductive airway tree at which this happens (23).
Table 1 shows the baseline values of all pertinent spirometry- and MBW-derived indexes in the equal-dose and double-dose groups, showing no significant difference between histamine and methacholine baseline (Wilcoxon signed-rank test, P> 0.1). Table 2 summarizes the differences induced by 6.52 μmol histamine or 6.67 μmol methacholine in the equal-dose group and by 6.52 μmol histamine or 13.3 μmol methacholine in the double-dose group. Note that impaired spirometry is reflected in FEV1, PEF, or FEF75 decreases (negative numbers in Table 2), whereas impaired ventilation heterogeneity is indicated byS cond and S acin increases and Curv decreases (see Theoretical background). Also indicated in Table 2 are the P values of significantly different changes after either histamine or methacholine. For instance, FEV1 was invariably less affected by methacholine than by histamine in both groups (P = 0.002 equal-dose group;P = 0.009 double-dose group). By contrast, VDanat decreased significantly by 10–15 ml after each final dose of either provocation agent, but no significant VDanat were obtained between histamine and methacholine in the same subjects.
Any parameter that showed a significantly different modification after methacholine with respect to histamine in Table 2 is also graphically represented in Figs. 2-6, where panels A and B represent data obtained from the equal-dose and double-dose groups, respectively. The spirometric indexes are also depicted for the intermediate steps throughout the provocation procedures. Figure 2 indeed shows consistently more marked FEV1 decreases with histamine than with methacholine for an equal cumulative dosage (Fig. 2) or when doubling the final methacholine dose with respect to the final histamine dose (Fig.2 B). The same pattern was observed for PEF (Fig.3). However, FEF75 was still affected more by histamine than by methacholine for equal dosage (Fig.4) but was indistinguishable between both agents for one more doubling methacholine dose (Fig. 4 B). Such contrasting behavior with respect to FEV1 is even more dramatic in the case of S cond (Fig.5). Although equal doses of histamine and methacholine only induced slightly smaller S condincreases for methacholine than for histamine (P = 0.04), a double dose of methacholine (Fig. 5 B) actually led to a larger S cond increase (P = 0.006) in the same subjects that produced a smaller FEV1decrease (Fig. 2 B). The S cond pattern was paralleled by that obtained for Curv (Fig.6).
The present study displays an apparent paradox by revealing a greater ventilation maldistribution for the bronchoconstriction agent (methacholine) that provokes the least deterioration of spirometry, at least in terms of FEV1. As in previous studies (22,26), the FEV1 behavior observed here (Fig. 2) would suggest that, in normal subjects, histamine is a more potent bronchoconstrictor than methacholine, even when the molar dose of methacholine is doubled with respect to histamine (Fig.2 B). Noteworthy for comparison with previous studies is that the maximal cumulative weight dose for histamine (2 mg) was intermediate between methacholine weight doses of the last step of the equal-dose group (1.6 mg) and that of the double-dose group (3.2 mg). Although histamine may be more potent than methacholine in the large airways, as indicated by the respective FEV1 and PEF decreases (Figs. 2-3), this may not be necessarily true for the small airways. Indeed, FEF75 decrease for methacholine tended to mimic FEF75 decrease seen for histamine (Fig. 4), with only a marginally significant difference for an equal molar dose of histamine and no difference at all for one more doubling dose of methacholine (Table 2). Such an observation could suggest that, in the small airways, the basic mechanisms of overall airway narrowing may be the same (direct effect at the level of the bronchial smooth muscle) but merely dose dependent (depending on the number distribution of either H1 or muscarinic receptors).
The ventilation distribution data provide an interesting new perspective on the subject, because its indexes are linked to heterogeneity of airway narrowing rather than an overall lumen decrease at a given lung depth. The two- to threefoldS cond increases and the absence ofS acin change after histamine or methacholine provocation (Table 2) indicate considerable parallel heterogeneity of airway narrowing in the conductive airway tree. An equal molar dose of histamine or methacholine (equal-dose group in Table 2) shows only marginal significance of the differences inS cond increase (P = 0.041) in the presence of highly significant differences in FEV1decline (P = 0.002). This could have suggested thatS cond is rather variable and not very specific in response to provoking agents that clearly constrict the large airways to a different extent (as evidenced by FEV1 and PEF behavior). However, a doubling dose of methacholine with respect to histamine (double-dose group; Table 2) reveals a clear-cut divergence between a larger S cond increase (P = 0.006) and a smaller FEV1 decline (P = 0.009) for methacholine. Clearly, the latter ventilation distribution and spirometric parameters reflect different phenomena occurring simultaneously.
The parallelism between respective S cond and Curv patterns seen across Figs. 5 and 6 suggests that the respectiveS cond changes for histamine and methacholine can be accounted for to a large extent by differences in inspired gas concentration (or specific ventilation) in addition to flow asynchrony patterns developed by the heterogeneity in airway narrowing. In this respect, the present data could explain earlier observations of dissociation between spirometric decline and ventilation-perfusion mismatch seen in mildly asthmatic subjects after methacholine (17). If, in the case of bronchoprovocation, the increased ventilation-perfusion mismatch can be accounted for mainly by ventilation heterogeneities, the present data predict a more marked dissociation between spirometry and ventilation-perfusion mismatch for methacholine than for histamine, at least in normal subjects.
On the basis of S cond or Curv, it is a priori not possible to locate the heterogeneity of airway narrowing to proximal or peripheral airways within the conductive airway tree (i.e., distinguish between large and small conductive airways). Alternatively, spirometry may not be able to adequately represent narrowing of the most peripheral airways in the conductive airway tree, i.e., the entrance to the silent zone of the lungs. It is nevertheless tempting to speculate on a possible link between the behavior of ventilation maldistribution and spirometry. Let us consider the double-dose group results, in which FEV1 decreases less (Fig. 2), FEF75 decreases to the same extent (Fig. 4), andS cond increases more (Fig. 5) after methacholine than after histamine. If the methacholine-inducedS cond increases originated in the large airways, this agent would need to have provoked a large heterogeneity in airway narrowing among a relatively small number of parallel airways while maintaining a milder overall airway narrowing than with histamine (as indicated by FEV1 and PEF). The alternative possibility seems more likely: methacholine-induced S condoriginates in the small airways, where it provokes a larger heterogeneity in airway narrowing among a relatively large number of parallel airways, for an overall airway narrowing similar to that obtained with histamine (as indicated by FEF75). We should also not exclude the possibility that S cond may have originated in conductive airways that are even smaller than those that can be reflected in FEF75. Under this assumption, FEF75 merely showed a transitional behavior between the large airways (reflected in FEV1 and PEF) and these smaller airways (reflected in S cond and Curv). Finally, we consider the possibility that heterogeneity may also affect the spirometric indexes, despite experimental data by McNamara et al. (13) showing that the forced expired flow-volume relationship fails to reflect interregional heterogeneity of alveolar pressure, as induced by histamine inhalation in dogs. If heterogeneity were a major contributor to FEV1, our observations would actually imply that the worst ventilation heterogeneity (as induced by methacholine) causes the least deterioration in FEV1.
The present data may give new impetus to the search for the mechanisms of histamine and methacholine action along the bronchial tree. TheS cond results strongly suggest that when characterizing any mechanism of airway narrowing (e.g., localization of receptors) or when visualizing airway narrowing itself [e.g., high-resolution computed tomography (HRCT)-derived airway lumen], one should not only look for number of receptors or for overall airway lumen decreases at different lung depths but should quantify the parallel heterogeneity in distribution of these receptors, or heterogeneity in constriction of airways, at any given lung depth as well. Because the differentiation between histamine and methacholine clearly occurs within the conductive lung zone, present HRCT resolution (12) is probably sufficient to quantify heterogeneity of airway narrowing. Such information could then be fed into models such as those proposed by Gillis and Lutchen (9) to estimate resulting distributions of specific ventilation, from whichS cond and Curv values can be computed for quantitative comparison with experimental data. Also, the similarity between S cond and Curv (i.e., the significant role of specific ventilation differences) suggests that differences between histamine and methacholine could be revealed by three-dimensional imaging of ventilation distribution. If the hypothesis holds that S cond originates at least in part from the smaller, but still conductive, airways, ventilation maldistribution will be operational between lung units as small as groups of acini, and resolution may be a limiting factor.
Sekisawa et al. (19) previously suggested that methacholine was operational mainly in the large airways, whereas histamine affected both large and small airways, on the basis of the respective changes of VDanat and respiratory resistance (and considering equal histamine and methacholine concentrations in terms of mg/ml). This conclusion has been challenged by other studies using HRCT in dogs (3) and humans (15), demonstrating at least the ability of methacholine to contract the small airways. In the dog study (3), methacholine was in fact seen to provoke a slightly greater airway narrowing than histamine throughout the airway tree. In contrast to Sekisawa et al. (19), we did not find any significant difference between VDanat decreases after either histamine or methacholine. In the present study with nonhyperresponsive normal subjects, VDanat changes were small (∼10 ml) and subtle differences may have disappeared in the measurement noise, although FRC and breathing frequency were well controlled and showed no significant intrasubject differences across all baseline and provocation stages.
In summary, the present study illustrates a clear dissociation between outcome of spirometry and ventilation heterogeneity in response to methacholine and histamine challenges in normal subjects. However, such observations are not incompatible if both overall airway narrowing and parallel heterogeneity are considered. The present data therefore suggest that future efforts toward the understanding of histamine and methacholine along the airway tree should be oriented toward quantification of both serial and parallel distribution of airway constriction or the underlying sources of constriction. Our data strongly suggest that these should be found in the conductive airway tree, and we speculate that the larger parallel heterogeneity in constriction seen for methacholine than for histamine may originate in the small conductive airways.
We thank Johan Goris from the Biotechnology Department of Akademisch Ziekenhuis, Vrije Universiteit Brussel, Belgium for technical support.
This study was funded by the Fund for Scientific Research-Flanders and the Federal Office for Scientific Affairs (program PRODEX).
Address for reprint requests and other correspondence: S. Verbanck, AZ-VUB, Consultatie Pneumologie, Laarbeeklaan 101, 1090 Brussels, Belgium (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2001 the American Physiological Society