|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Pulmonary-Critical Care Medicine Branch (Dr. Taveira-DaSilva, Ms. Hedin, Ms. Hathaway, and Dr. Moss) and Office of Biostatistics Research (Dr. Stylianou), National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD.
Correspondence to: Joel Moss, MD, PhD, Pulmonary-Critical Care Medicine Branch, National Heart, Lung, and Blood Institute, National Institutes of Health, Building 10, Room 6D-05, MSC 1590, Bethesda, MD 20892-1590; e-mail: mossj{at}nhlbi.nih.gov
| Abstract |
|---|
|
|
|---|
Design: Retrospective study.
Setting: National Institutes of Health, Bethesda, MD.
Design and subjects: The study population comprised 348 patients with LAM participating in a longitudinal research protocol. Declines in diffusion capacity of the lung for carbon monoxide (DLCO) and FEV1 were measured in 275 patients observed for approximately 4 years. The declines in DLCO and FEV1 of patients treated with progesterone, po (n = 67) or IM (n = 72), were compared with those of untreated patients (n = 136).
Measurements and results: Overall yearly rates of decline in DLCO and FEV1 were 2.4 ± 0.4% predicted (0.69 ± 0.07 mL/min/mm Hg) and 1.7 ± 0.4% predicted (75 ± 9 mL), respectively (mean ± SEM). The most significant predictors of functional decline were initial lung function and age. After adjusting for initial FEV1, age, and duration of disease, patients treated with IM progesterone tended to have lower rates of decline in FEV1 than patients treated po (1.9 ± 0.6% predicted vs 3.2 ± 0.8% predicted, respectively; p = 0.081). However, there was no significant difference in rates of decline in FEV1 between patients treated with IM progesterone and untreated patients (1.9 ± 0.6% predicted vs 0.8 ± 0.5% predicted, respectively; p = 0.520), and patients treated with po progesterone and untreated patients (3.2 ± 0.8% predicted vs 0.8 ± 0.5% predicted, respectively; p = 0.064). After adjusting for initial DLCO, rates of decline in DLCO were significantly higher in patients treated with po progesterone (3.6 ± 0.7% predicted, p = 0.002) and IM progesterone (2.8 ± 0.5% predicted, p = 0.022) than in untreated patients (1.6 ± 0.6% predicted).
Conclusions: Within the limitations of a retrospective study, our data suggest that progesterone therapy does not slow the decline in lung function in LAM.
Key Words: interstitial lung diseases progesterone respiratory function
| Introduction |
|---|
|
|
|---|
LAM occurs sporadically in patients with no evidence of genetic disease and in approximately one third of women with tuberous sclerosis complex (TSC).5678 Based on the prevalence of TSC,9 it is estimated that there may be as many as 7,500 TSC patients with LAM in the United States.567 Sporadic LAM, however, which is not associated with a TSC phenotype, is a relatively uncommon disease with a clinical prevalence estimated at 2 to 6 per million women.1234
Pulmonary function abnormalities in LAM consist of gas exchange abnormalities, characterized by a decreased diffusion capacity of the lung for carbon monoxide (DLCO) and airflow obstruction.1011 In many patients, progressive lung damage occurs and functional impairment is severe, leading to incapacitation and complete disability, requiring long-term oxygen therapy and lung transplantation, or causing death. The rate of progression of disease, however, is variable, and some patients have a long-term course lasting > 20 years.1011
There is some evidence suggesting that LAM may be influenced by hormonal factors. Indeed, not only does LAM affect primarily women,1234 but the disease appears to progress during pregnancy,1213 or following the administration of estrogens.141516 In addition, there is evidence for the colocalization of estrogen and progesterone receptors in LAM cells.17181920 These findings have prompted the use of hormonal therapy, especially with progesterone, for the treatment of LAM. However, no studies have been undertaken in large populations of patients with LAM to evaluate the rate of decline of lung function in LAM and to determine whether this decline is affected by treatment with progesterone.
The aim of this study was to determine whether progesterone therapy slowed the decline in lung function in LAM. To this end, we evaluated our population of 348 patients with LAM for progesterone use. We were able to determine the rates of decline in lung function in 275 patients, and compared the rates of patients treated with progesterone with those of patients who did not receive progesterone.
| Materials and Methods |
|---|
|
|
|---|
The decision to initiate progesterone therapy, and the choice of route of administration, was not part of the NHLBI protocol and was made, in the majority of the patients, independently, by the patients physicians. Patients were evaluated approximately every 6 months (range, 5 to 7 months). Compliance with progesterone therapy was monitored by interviewing the patient at the time of each visit. Patients were considered to have reached menopause when menopause occurred naturally (low estradiol levels and elevated follicle-stimulating hormone levels) or was surgically induced (bilateral oophorectomy, without hormonal replacement therapy).
Pulmonary Function Tests
Lung volumes, flow rates, and DLCO were measured using a computerized system (Collins Gold Standard Plus; Warren E. Collins; Braintree, MA; or Master Screen PFT, Erich Jaeger; Wuerzburg, Germany) according to American Thoracic Society standards.212223
Statistical Analysis
The available data set contained multiple pulmonary function measurements for each of the 275 patients. We summarized the information from each subject by using the estimated yearly rate of decline (slope), calculated from a linear regression using the percentage of predicted DLCO and FEV1 as the response variables and the time of each test as the independent variable, considering the first test as time zero. No covariate adjustment was used in deriving the rate of change in lung function, since the percentage of predicted lung function is already covariate adjusted. The effect of progesterone treatment on the yearly rate of lung function decline was tested using a two-sided t test. In addition, we ran univariate and multivariate regression analyses to identify factors related to the rate of decline in pulmonary function. Among the explanatory variables considered were age; duration of disease, measured as time since diagnosis and time since first LAM-related symptoms; menopausal state; treatment or no treatment with progesterone; route of administration of progesterone therapy; initial DLCO or FEV1; and time under observation.
The three groups of patients were as follows: patients with no progesterone administered, patients with progesterone administered po, and patients with progesterone administered IM. A one-way analysis of variance with a Fisher adjustment was employed to compare patient groups.
| Results |
|---|
|
|
|---|
Pulmonary Function Tests and Rate of Decline in Lung Function
The average number of tests per patient was 6.0 ± 0.2 (range, 2 to 14). Initial pulmonary function test results are shown in Table 1
. After an average follow-up time of, respectively, 3.9 ± 0.2 years and 4.0 ± 0.2 years, DLCO had declined from 15.9 ± 0.4 mL/min/mm Hg (72.7 ± 1.6% predicted) to 13.5 ± 0.3 mL/min/mm Hg (65.2 ± 1.5% predicted), and FEV1 from 2.09 ± 0.04 L (75.2 ± 1.4% predicted) to 1.83 ± 0.05 L (69.5 ± 0.6% predicted). The mean yearly declines in DLCO and FEV1 were 2.4 ± 0.4% predicted (0.69 ± 0.07 mL/min/mm Hg) and 1.7 ± 0.4% predicted (75.0 ± 9.0 mL), respectively.
|
|
Predictors of Decline in DLCO
Of the potential predicting variables (initial DLCO and FEV1, age, duration of disease, time under observation, menopausal state, treatment or no treatment with progesterone, and route of administration of progesterone), only initial DLCO (r = 0.25, p < 0.0001) and FEV1 (r = 0.12, p = 0.039) were significantly correlated with the decline in DLCO. Patients with higher initial DLCO and FEV1, ie, with less severe disease, tended to have a greater rate of decline in DLCO.
Predictors of Decline in FEV1
The rate of decline in FEV1 was negatively correlated with the initial DLCO (r = 0.17, p = 0.004) and age (r = 0.22, p = 0.0003). That is, older patients and patients with higher DLCO and FEV1 had a lower rate of decline in FEV1. Follow-up time, ie, the time under observation, was not correlated with the yearly rate of decline in DLCO (p = 0.257) or FEV1 (p = 0.835).
In a multivariate regression analysis, we confirmed that initial DLCO (p < 0.001) was a significant predictor of the rate of decline in DLCO. Both initial DLCO (p < 0.001) and the patient age (p < 0.001) were significant predictors of the rate of decline in FEV1.
Duration of disease was not a significant predictor of the decline in DLCO and FEV1 except in the group of patients who were treated with progesterone. Using a cluster analysis, where variables are clustered into groups based on their intercorrelations, we confirmed that length of disease falls into the same variable cluster as time of follow-up, which is a better predictor of the rate of lung function change.
Decline in DLCO and FEV1 in Patients Treated With and Without Progesterone: Adjusted Analysis
Of the 275 patients, 139 were treated with progesterone (67 po and 72 IM). The mean duration of progesterone therapy for all was 56 ± 4 months, and the average monthly dose was 580 ± 37 mg. The mean duration of therapy for patients treated with oral progesterone was 48 ± 5 months, and the average monthly dose was 714 ± 68 mg. For patients treated with IM progesterone, the mean duration of therapy was 63 ± 6 months and the average monthly dose was 462 ± 29 mg. The monthly dose of progesterone was significantly lower (p < 0.001) in the IM group, but the cumulative totals of administered drug were not significantly different: 40,600 ± 9,810 (po) vs 30,440 ± 3,660 mg (IM), p = 0.313.
Differences among the three groups of patients in age, initial lung function, years since diagnosis, and observation time (follow-up time) are shown in Table 3 . Progesterone-treated patients had lower initial DLCO and FEV1 than did nontreated patients, and they also tended to have a longer time of follow-up. Additionally, patients treated with IM progesterone were significantly younger. Consequently, to test for a treatment effect adjusted for baseline differences between the three groups, a multivariate regression model was used, adjusting for one or more of these significant variables: initial lung function, age, duration of disease (where significant), and route of therapy.
|
No Progesterone vs po Progesterone
The rate of decline of percentage of predicted DLCO, after adjusting for initial DLCO, was significantly lower (p = 0.002) in the untreated group than in the group treated with po progesterone (Table 3, Fig 1
). After adjusting for initial FEV1 and age, the rate of decline of percentage of predicted FEV1 in the untreated group was lower than that of patients treated with po progesterone, but the difference was not statistically significant (p = 0.064) [Table 3, Fig 1].
|
IM vs po Progesterone
In this analysis, duration of disease, initial FEV1, and age were significant predictors of rates of decline in DLCO and FEV1. Patients with greater duration of disease had a lower rate of decline in lung function. After adjusting for age, duration of disease, and initial lung function, we found a tendency toward a lower rate of decline of percentage of predicted FEV1 in the IM progesterone group (1.9 ± 0.6% predicted vs 3.2 ± 0.8% predicted). However, the difference was not statistically significant (p = 0.081). There was no significant difference between these two groups in the rates of decline of percentage of predicted DLCO (2.8 ± 0.5% vs 3.6 ± 0.9%, p = 0.800) [Table 3, Fig 1].
| Discussion |
|---|
|
|
|---|
Postmenopausal patients tended to have lower rates of decline in FEV1 than premenopausal patients, although the difference in rates of FEV1 decline was not statistically significant. This was seen despite the fact that postmenopausal patients had a significantly lower DLCO (p = 0.037) than premenopausal patients that, based on our data, should have been associated with a higher rate of decline in FEV1. Conversely, postmenopausal patients were significantly older than premenopausal patients, and the rate of decline in FEV1 decreases with age. Thus, age and menopause, or a combination of both factors, could have accounted for a lower rate of decline in FEV1 in postmenopausal patients.
Our data demonstrate the difficulties encountered in evaluating retrospectively the effects of therapeutic interventions on lung function in heterogeneous patient populations. Although the multivariate analysis attempted to adjust for factors found to correlate with decline in lung function, it is still possible that treated and untreated patient populations may have been substantially different in the actual rates of disease progression. It is also conceivable that patients who appeared to be sicker were more likely to be treated by their physicians, whereas those with milder disease chose not to receive or were not offered progesterone therapy.
The efficacy of hormonal therapy with progesterone in slowing the progression of lung disease in LAM has not been established. Reports of success, based on few patients, or without objective data, are matched by reports2526272829303132333435 of therapeutic failures. In one study36 of premenopausal patients, the yearly decline in DLCO was less in 16 patients treated with progesterone than in 13 untreated patients. However, the initial DLCO was lower in the treated group, which, according to our data, was associated with a lower rate of decline in DLCO. In five patients for whom pretreatment and posttreatment decline rates for FEV1 and DLCO were available, it was found that progesterone decreased the rate of decline in lung function. In this analysis, it was assumed that there was a constant rate of decline in lung function that was diminished by progesterone. This assumption is probably incorrect because our study suggests that the rate of decline in lung function may not be constant over time, ie, as the DLCO declines, the rate of decline in DLCO decreases, and the rate of decline in FEV1 increases. Therefore, any comparison of rates of decline in lung function between LAM patient groups needs to adjust for initial function to ensure that the initial severity and rate of progression of disease are reasonably similar in all groups.
After adjusting for initial lung function and age, we found no benefit of progesterone in decreasing the rate of decline in lung function. We observed a trend toward a lower rate of decline in FEV1 in patients treated with IM progesterone than in those treated with po progesterone. Presumably, the lower bioavailability of po progesterone3738394041 could account for the tendency for higher rates of decline in FEV1 in patients treated po than in those treated by the IM route. However, the finding that patients treated with progesterone po had a greater rate of decline in FEV1 than untreated patients cannot be explained by a lower drug availability. Conversely, the possibility that po progesterone might actually produce adverse effects on FEV1 cannot be excluded. In addition, since treatment with IM progesterone was not significantly better than no treatment, it appears that treatment with this hormone by either route of administration offers no benefit to patients with LAM.
Our data raise questions about the justification for committing a patient with LAM to prolonged treatment of a chronic disease with an agent of questionable benefit and significant side effects.424344454647484950 However, because this was a retrospective study and the patient populations were heterogeneous, our findings have to be interpreted with caution. The question of whether treatment with progesterone is of benefit to patients with LAM may never be completely answered unless a prospective therapeutic trial comparing IM progesterone with a placebo is undertaken. Our observations should provide useful information for designing therapeutic trials in LAM.
| Acknowledgements |
|---|
| Footnotes |
|---|
Supported by the NHLBI, Division of Intramural Research.
Ms. Xiaoling Chen was supported in part by a grant from the LAM Foundation.
Received for publication January 12, 2004. Accepted for publication July 23, 2004.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
F. X. McCormack Lymphangioleiomyomatosis: A Clinical Update Chest, February 1, 2008; 133(2): 507 - 516. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Harari, R. Cassandro, J. Chiodini, A. M. Taveira-DaSilva, and J. Moss Effect of a Gonadotrophin-Releasing Hormone Analogue on Lung Function in Lymphangioleiomyomatosis Chest, February 1, 2008; 133(2): 448 - 454. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. H. Ryu, J. Moss, G. J. Beck, J.-C. Lee, K. K. Brown, J. T. Chapman, G. A. Finlay, E. J. Olson, S. J. Ruoss, J. R. Maurer, et al. The NHLBI Lymphangioleiomyomatosis Registry: Characteristics of 230 Patients at Enrollment Am. J. Respir. Crit. Care Med., January 1, 2006; 173(1): 105 - 111. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. F. Abbott, M. L. Rosado-de-Christenson, A. A. Frazier, T. J. Franks, R. D. Pugatch, and J. R. Galvin From the Archives of the AFIP: Lymphangioleiomyomatosis: Radiologic-Pathologic Correlation RadioGraphics, May 1, 2005; 25(3): 803 - 828. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |