INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Enlargement of visceral organs such as the lungs, heart, and kidneys is a well recognized manifestation of acromegaly (1). Although the earliest reports suggested that large lung volume was restricted to some males (1), it has subsequently been described in female patients as well (2). Since the increased lung volume in acromegalic patients is not related to hyperinflation (1) or to increased inspiratory muscle strength (3), it has been accepted that the excess of growth hormone (GH) in acromegaly induces growth of the lungs (1).

However, the mechanism by which increased lung growth occurs in acromegaly is not clear. Classically, it had been reported that lung compliance was increased in acromegalic patients, whereas the lung elastic recoil, the diffusing capacity, and the diffusing capacity per unit of alveolar volume (VA) were normal (1, 4). These findings suggested that lung growth in acromegaly results from an increase in alveolar size resulting from an alteration of the elastic properties of the lungs (1). Surprisingly, in a recent study, Donnelly and colleagues (5) found that acromegalic patients had a similar pulmonary distensibility, K (an index of alveolar size) (6), and diffusing capacity to those of control subjects. However, diffusing capacity per unit of VA was lower in acromegalic patients than in normal subjects. In view of these results, the authors deduced that alveolar multiplication was the growth mechanism in the acromegalic lung. Differences between the studies in which these findings were made could be attributed to some methodologic problems. In some studies, there were no control groups (1, 4), whereas in another study (5), the control group was not well matched with the acromegalic group.

Another aspect to consider is the influence of disease activity on lung growth. Previous studies had reported that suppression of GH hypersecretion in acromegaly was followed by a reduction in left ventricular mass (7) and by morphologic changes in the wall of the upper airway (8). Thus, it may be speculated that if lung growth in acromegalic patients is due to an alteration of the elastic properties of the lungs, treatment might induce some change in alveolar size. Meanwhile, a modification of the number of alveoli caused by a decrease in GH should be more improbable. The aim of this study was to compare pulmonary distensibility, K, and diffusing capacity in a group of patients with active acromegaly with those in a group of age-, sex-, height- and weight-matched control subjects. We also evaluated the response of these parameters to the suppression of GH hypersecretion.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We studied six males and five females with active acromegaly and 11 control subjects. The patients and controls were matched for sex, age, smoking habit, weight, and height. The control subjects were judged to be healthy on the basis of history, physical examination, electrocardiogram, basal spirometry, and chest radiography.

Patients were treated according to the clinician's criterion with surgery or somatostatin analogs. Acromegaly was considered to be inactive when baseline GH levels were < 5 µg/L and GH concentrations were < 2 µg/L after glucose loading (9).

All subjects were informed of the experimental requirements of the study before attendance, and signed consent forms.

Measurements

Plasma GH and insulin-like growth factor (IGF)-1 concentrations were determined with automated immunoenzymatic assay (AIA 1200; Tosoh Corporation, Tokyo, Japan) and radioimmunoassay kits (Nichols Institute, San Juan Capistrano, CA) respectively. Arterial blood gases were measured during rest. Spirometry and plethysmography were performed as previously described (10).

Carbon monoxide diffusing capacity (DL_CO) was measured by the single-breath method of Ogilvie and colleagues (11), using the MasterLab 3.2 system (Jaeger, Würzburg, Germany). Measurements were repeated after a 5-min interval. At least two acceptable tests (DL_CO difference < 5% and difference in inspired volume > 90% of VC) were performed, and the results were expressed as mean values of the two closest measurements. DL_CO was corrected for hemoglobin levels (12).

Maximum inspiratory (PImax) and expiratory (PEmax) pressures were measured in a standard procedure (13), using a differential pressure transducer (M-163; Sibelmed, Barcelona, Spain). The maneuvers were repeated until three measurements, sustained for at least 3 s each and with < 5% variability, were recorded. The highest value obtained was used for analysis.

Lung elastic recoil pressure was obtained from quasistatic expiratory pressure-volume (P-V) curves. Transpulmonary pressure (PL) was measured as the difference between mouth and esophageal pressure, using the esophageal balloon method (14). A pressure transducer (Jaeger) was used to measure PL, and volume was recorded with a pneumotachograph (Model 276; Jaeger). Expiratory P-V curves were recorded between TLC and FRC (measured with plethysmography) with expiratory flow < 0.5 L/s (14).

Quasistatic lung compliance (CLqst) was obtained by dividing the volume change of 0.5 L by the change in PL between FRC and FRC + 0.5 L. Specific compliance (SCLqst) and PL at 90% and 100% of actual TLC were also computed (14).

All P-V points were fitted with an iterative least-mean-squares regression, according to the exponential equation V=ABe^KP (15). The quality of fit of the exponential equation to the data for each subject was assessed from the standard errors of the coefficients (K, A, and B) and the reduction of the original variance (r²). An arbitrary r² value of 0.96 was chosen, and values below the scatter of the points around the line were considered unacceptable (16).

Statistical Analysis

Results are expressed as mean ± SD. Mann-Whitney U test was used to evaluate differences between acromegalic patients before treatment and normal subjects. Likewise, Wilcoxon's test was used to compare acromegalic patients before and after treatment. Association between variables was computed by means of Spearman's correlation coefficient. A value of p < 0.05 was accepted as the minimum level of statistical significance (17).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The clinical characteristics and endocrine status of acromegalic patients are presented in Table 1. As shown by the increased levels of IGF-1 and GH, all patients were in an active phase of disease. Body mass index (BMI) ranged from 23.6 to 35.8 kg/m². Five patients were obese (BMI > 30 kg/m²). In control subjects, levels of GH and IGF-1 were 0.63 ± 0.91 (mean ± SD) µg/L and 207 ± 50 µg/L, respectively.

The results of the pulmonary function tests are presented in Tables 234. Acromegalic patients had a greater FRC and TLC than did normal subjects. No significant differences were found between the acromegalic patients and normal subjects in arterial blood gases, DL_CO, or PImax. Conversely, a relevant difference in lung distensibility was observed in the acromegalic versus the normal subjects. Acromegalic subjects had an increased CLqst and SCLqst, and reduced PL at 90% and 100% of actual TLC, as compared with control subjects.

In all subjects, the points on the P-V curves fitted the exponential equation with r² > 0.96 (mean r² ± SD: 0.98 ± 0.01). The shape constant, K, was significantly larger in patients with active acromegaly than in control subjects. No changes in the volume extrapolated to infinite transpulmonary pressure, A, or in the index of position of the P-V curve, B/A, were noted. The mean intrasubject standard error (SE) of K was 0.22 ± 0.14 in patients with active acromegaly, 0.24 ± 0.03 in those with inactive acromegaly, and 0.21 ± 0.09 in normal subjects. The SEs of the coefficients A and B were 0.24 ± 0.03 and 0.18 ± 0.06 in patients with active acromegaly, 0.23 ± 0.05 and 0.23 ± 0.07 in those with inactive acromegaly, and 0.21 ± 0.02 and 0.24 ± 0.03 in control subjects, respectively.

After 5 ± 3 (range: 3 to 11) mo of treatment, all patients were judged to have inactive acromegaly (mean GH: 3.4 ± 2.4 µg/L; mean IGF-1: 515 ± 226 µg/L), and no significant changes in arterial blood gases, DL_CO, or maximal respiratory pressures were observed. FRC, TLC, CLqst, SCLqst, and the K coefficient (Figures 1 and 2) were significantly decreased when compared with those before treatment. After treatment, maximum transpulmonary pressure (PLmax) and transpulmonary pressure at 90% of TLC (PL90) increased; and coefficients A and B/A were unchanged. There were no significant differences in lung distensibility between patients with inactive acromegaly and control subjects (Table 4).

View larger version (18K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 1.   Lung P-V curves in acromegalic patients before (left) and after GH hypersecretion suppression (right). The closed circles represent PL at FRC in each patient.

View larger version (20K): [in this window] [in a new window]   View larger version (22K): [in this window] [in a new window]   Figure 2.   Individual CLqst (A) and shape constant of the pressure-volume curve (K coefficient) (B) during active and inactive acromegaly. Horizontal bars represent mean values.

Duration of disease at the time of diagnosis was not correlated with TLC (r = 0.055, p = 0.872), FRC (r = 0.405, p = 0.217), CLqst (r = 0.189, p = 0.579), SCLqst (r = 0.050, p = 0.883), PLmax (r = 0.145, p = 0.671), or the K coefficient (r = 0.275, p = 0.413). Similarly, no significant correlation was found between baseline GH levels and TLC (r = 0.387, p = 0.239), FRC (r = 0.544, p = 0.053), CLqst (r = 0.110, p = 0.748), SCLqst (r = 0.223, p = 0.509), PLmax (r = 0.244, p = 0.469), or the K coefficient (r = 0.328, p = 0.325).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study provides evidence that: (1) patients with active acromegaly have a greater pulmonary distensibility and lower PL than do normal subjects; (2) the diffusing capacity and diffusing capacity per unit of alveolar volume do not differ between acromegalic subjects and controls; and (3) the suppression of GH hypersecretion reduces lung volumes and lung distensibility in patients with acromegaly.

Previous studies of lung distensibility in acromegaly gave contradictory results. Brody and colleagues (1) reported that lung compliance was high and PL was normal at TLC and FRC in three acromegalic males. In contrast with this, as well as with the present study, Donnelly and colleagues (5) found no significant difference in either pulmonary distensibility, determined with the exponential constant K, or elastic recoil in acromegalic patients versus normal subjects. It is possible that the differences between the given findings were the result of bias in selection of the control groups, because the acromegalic patients studied by Donnelly and colleagues were of a significantly greater age and weight than were the healthy subjects. In their series, mean BMI was 30.3 kg/m² in acromegalic males and 28.2 kg/m² in acromegalic females. Classical reports demonstrate that age and weight affect the P-V curve (5, 18). Lung compliance is decreased by approximately 25% in simple obesity, probably due to the related increase in pulmonary blood volume and increased closure of dependent airways (18). Moreover, the use of predicted percentage did not completely resolve this bias because Colebatch and coworkers' normal values for K were based on age and height but did not consider weight (15).

The relationship between alveolar distensibility and alveolar size has been fully studied. The exponential constant K has been shown to be related to the mean linear intercept (Lm), a morphometric estimate of the mean size of alveolar air spaces at TLC in rats, cats, and dogs (6), and in humans, including normal subjects, smokers (19), and persons with emphysema (20). Elastic recoil is dependent on alveolar size because surface tension is inversely related to Lm and directly related to the alveolar surface-to-volume ratio. However, although surface forces have the predominant influence on lung distensibility (6), tissue elastic properties could also contribute to lung distensibility because tissue elements provide a degree of structural rigidity for alveolar septa, and facilitate the transmission of forces between contiguous air spaces. Administration of recombinant human GH induces lipolysis, protein anabolism, and muscle growth (21). Studies in vitro and in vivo have shown that GH increases the synthesis of type I collagen fibers, and especially mucopolysaccharides, in the lung (21). There is some controversy about the sites of direct GH action. GH receptor/GH binding proteins are distributed in the epithelium and smooth muscle, or in cells of the bronchoalveolar tree, including type I and II pneumocytes, in rats (22). Moreover, GH seems to play a role in compensatory postpneumonectomy lung growth in rats (23). However, in human lungs, no binding sites for GH have been found until now, suggesting that this hormone does not play a direct physiologic role in human lung growth and development (24). On the other hand, the expression of IGFs, IGF receptor, and IGF binding proteins in human lungs suggests that IGFs have a role, possibly GH-induced, in the growth and development of human lungs (25). This could be the explanation for our inability to demonstrate any correlation between the level of GH and K or lung volumes. Other authors (4, 26) also reported finding no relationship between GH level and lung volumes.

Our findings of a normal diffusing capacity and diffusing capacity per unit of VA in acromegalic patients are supported by previous reports. Brody and colleagues (1) measured diffusing capacity in six men and four women with acromegaly and found normal values in all of them. Evans and colleagues (4) described pulmonary function in 12 females and eight males with acromegaly. In both sexes, the diffusing capacity and the ratio of diffusing capacity to alveolar volume (DL/VA) were normal with respect to their predicted values. More recently, Trotman-Dickenson and colleagues (27) reported normal percentages of diffusing capacity and DL/VA in acromegalic patients. The lack of increase in DL_CO found in our patients could be explained by a decrease in the surface-to-volume ratio of the lung. Moreover, this could in part be compensated by an increase in blood volume, which these patients probably have. The decrease in both parameters with treatment could leave DL_CO unchanged.

Donnelly and colleagues (5) attributed the decrease that they found in DL/VA to an increase in the unperfused capillary bed caused by the greater perfusion distances and the lower pulmonary capillary blood volume in the acromegalic lung. Correction of DL_CO for variation in hemoglobin concentration was not made by these authors, and hemoglobin concentration was measured only in four women. Our results and previous reports are not consistent with this hypothesis. In the present study, DL/VA was similar among acromegalic and control subjects. In addition, moderate hypoxemia was not detected in any acromegalic patient (range: 9.5 to 13.7 kPa), nor were other gas alterations detected that suggested ventilation/perfusion mismatching. The finding of normal DL/VA values in trained swimmers with large lungs (28) is also explained with difficulty by an increase in the unperfused capillary bed caused by greater perfusion distances. The most striking finding in our study was that suppression of GH hypersecretion in patients with active acromegaly reduces lung volumes and lung distensibility without modifying diffusion capacity. The changes in lung elastic properties of our patients suggest either changes in surface forces or an increase in the strength of tissue tethering after treatment. We believe that these changes are difficult to justify on the basis of a decreased alveolar number. Certainly the large lung volumes of patients with active acromegaly, which are partly normalized after GH suppression, might be attributed to a process of alveolar hypertrophy rather than to hyperplasia.

Other effects of suppressing GH hypersecretion in acromegalic patients have been described. Somatostatin analogs reduce upper airway soft tissue volume in these patients (8). It has also been shown that octreotide induces a significant and rapid (7-d) reduction in left ventricular mass in acromegalic patients with left ventricular hypertrophy (7). An opposing situation that confirms the reversibility of GH-induced lung disorders is GH deficiency. Adult patients with GH deficiency diagnosed in childhood had lower FVC and TLC than did matched control subjects, but did not show any significant differences in RV and DL_CO (29). After 12 mo of GH replacement therapy, patients with childhood-onset GH deficiency showed a significant improvement in lung volume and maximal respiratory mouth pressure, but did not experience any significant change in DL_CO (30).

In summary, the increased lung volumes and lung distensibility with normal diffusion capacity demonstrated in patients with active acromegaly are partly reversible after suppression of GH hypersecretion, suggesting that lung growth in acromegaly results from an increase in alveolar size rather than from an increase in the number of alveoli.