SEARCH

SEARCH BY CITATION

Keywords:

  • Systemic sclerosis;
  • Forced vital capacity;
  • Pulmonary function testing;
  • Lung involvement;
  • Diffusion capacity

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Objective

To determine the ability of initial forced vital capacity (FVC) of patients with scleroderma to predict subsequent pulmonary function deterioration.

Methods

Data on 78 patients with scleroderma were retrospectively collected and analyzed. FVC (percent predicted), diffusing capacity for carbon monoxide (percent predicted), and various clinical and laboratory parameters were recorded. Pulmonary function decline (outcome) was defined as at least a 15-point sustained decrease in FVC (percent predicted). Kaplan-Meier analyses were performed separately for 60 patients initially assessed within the first 3 years from disease onset (group A) and 16 patients whose FVC values in the fourth or fifth year from disease onset were ascribed as baseline measurements (group B).

Results

Based on baseline FVC, patients in each group were categorized into those with normal FVC (≥80% predicted) and those with decreased FVC (<80% predicted). In group A, the percent-predicted FVC of 89% of patients with normal initial FVC and of 75% of patients with reduced baseline FVC did not decrease by ≥15 points at 5 years (log rank P = 0.04). Four patients with decreased baseline FVC developed respiratory failure (FVC <50% predicted) versus none with normal initial FVC. Analysis of group B showed no difference between patients with normal baseline FVC and those with decreased FVC in the ability to further predict pulmonary function decline (log rank P = 0.13). Clinical and laboratory parameters (age, male sex, baseline diffusion capacity, anti–topoisomerase I, or duration of Raynaud's phenomenon preceding skin manifestations) were not associated with pulmonary function decline.

Conclusion

Measured within the first 3 years from disease onset, baseline FVC (percent predicted) may predict deterioration of pulmonary function in patients with scleroderma. Patients with normal pulmonary function at initial assessment are at low risk to develop considerable impairment of pulmonary function.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Pulmonary involvement is a prominent feature in systemic sclerosis and a significant cause of morbidity and mortality (1–5). The pathologic abnormalities include pulmonary fibrosis, which occurs in ∼80% of patients, and pulmonary hypertension (6). Histologically, the majority of patients with scleroderma with parenchymal disease have nonspecific interstitial pneumonia (NSIP), whereas a smaller portion have usual interstitial pneumonia (7–9). From a clinical standpoint, these 2 types of interstitial fibrosis have a different prognosis, with NSIP having a longer course and a relatively lesser decline of pulmonary function (10–12).

Sequential measurements of pulmonary function in patients with scleroderma have shown a remarkable variability in the pulmonary function decline, ranging from an indolent disease with a fairly constant pulmonary function to a rapidly deteriorating condition with considerable loss in both forced vital capacity (FVC) and diffusing capacity for carbon monoxide (DLCO) and ultimately development of respiratory failure (6, 13–15). In previous studies investigating the ability of various parameters to accurately predict pulmonary function loss in these patients (13, 16, 17), no consistent predictor of pulmonary function loss emerged. Male sex (18, 19), black race (15, 18, 20), presence of anti–topoisomerase I (anti–topo I) antibody (18, 20, 21), and severity of Raynaud's phenomenon (17) were all found to be associated with greater function loss.

The ability of initial pulmonary function to predict subsequent pulmonary function loss in patients with scleroderma has received particular attention. Baseline FVC was found to be a strong predictor of subsequent pulmonary function decline in some studies (6, 15) but not all studies (13, 16, 17, 22). It is likely that these disparate findings are in part due to methodologic differences among studies and heterogeneity in patient populations; there were also differences in the length of observation period (6, 14–16), disease duration at the time of first pulmonary function measurement (15, 16), and size of the population studied (6, 13–17, 23).

In this study, we reexamined the ability of baseline pulmonary function to predict further decline in pulmonary function in a large cohort of patients with scleroderma, paying particular attention to analyzing patients with similar lung involvement. For this purpose, we limited our analysis to patients with only parenchymal involvement and excluded patients with isolated pulmonary hypertension. Furthermore, to study the effects of timing of initial pulmonary function measurement relative to disease onset on the ability of baseline FVC to predict further pulmonary function loss, we carried out separate analyses using baseline FVC measurements obtained either relatively early (≤3 years) or late (>3 years) from disease onset.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Patients with scleroderma attending the outpatient University Rheumatogy Clinic at the Laiko University Hospital in Athens, Greece from 1995 to 2005 were identified, and their charts were retrospectively reviewed. All patients were diagnosed with scleroderma according to standard classification criteria (16, 24). Patients were included in the study if they were regularly followed up in our outpatient clinic for at least 3 years after disease onset and had serial pulmonary function tests at least twice during the study period. Parenchymal involvement was assessed by chest computed tomography; patients with isolated pulmonary hypertension as determined by cardiac echocardiography and no parenchymal disease were excluded. Patients with a coexisting cardiac failure or neuromuscular disease were also excluded.

For all patients, we recorded FVC (percent predicted), single breath DLCO (percent predicted), duration of Raynaud's phenomenon prior to disease onset, and presence of autoantibodies. Spirometry was performed according to standard techniques, using a spirometer (Masterlab; Jaeger, Wurzburg, Germany). Lung DLCO was determined with the single breath method.

On the basis of baseline FVC (percent predicted), patients were grouped into those with ≥80% predicted and those with <80% predicted FVC. The data were analyzed with the Kaplan-Meier survival method, with plot construction and estimation of log rank tests. The primary outcome in the analysis was a sustained decrease in percent-predicted FVC of at least ≥15 points. Separate Kaplan-Meier analyses were carried out for patients with initial FVC values obtained within 3 years of disease onset and for a second patient group whose FVC measurements taken after the first 3 years from disease onset were considered baseline values. Disease onset was defined as the appearance of skin manifestations other than Raynaud's phenomenon. Baseline DLCO (dichotomized as ≥70% and <70%), age (dichotomized as ≥45 years and <45 years), male sex, positive anti–topo I antibody, and duration of Raynaud's phenomenon preceding skin manifestations (dichotomized as ≥7 years and <7 years) were also examined as possible independent predictors of pulmonary function loss. All analyses were performed with commercially available statistical software (STATA version 8; Stata, College Station, TX). Data are presented as the mean ± SD; a P value less than or equal to 0.05 was considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

More than 240 charts of patients with scleroderma were initially reviewed; of those, 78 had serial measurements of pulmonary function over a range of years and were selected for analysis. The clinical and laboratory characteristics of these 78 patients are shown in Table 1. Of the 78 patients, 60 (group A) underwent their first pulmonary function measurement within a 3-year period from disease onset; 16 of the 78 patients (group B) were also included for separate analysis in which pulmonary function measurements made during the fourth or fifth year from disease onset were regarded as baseline measurements.

Table 1. Clinical and laboratory characteristics of scleroderma patients (n = 78)*
  • *

    Values are the percentage unless stated otherwise. FVC = forced vital capacity; PFT = pulmonary function testing; Anti–topo I = anti–topoisomerase I.

Age at diagnosis, mean ± SD years45.9 ± 13.5
Sex 
 Male15.4
 Female84.6
Smokers11.3
Disease duration, mean ± SD years9.97 ± 4.8
Time of first FVC measurement  from disease onset, mean ± SD  years1 ± 0.9
PFT measurements, median (range)4 (2–9)
Anti–topo I positive83.12

Among the group A patients, 28 (46.7%) of 60 patients had a normal baseline FVC (mean ± SD percent predicted 95.19 ± 10.11) and the remaining 32 patients (53.3%) had a decreased FVC (mean ± SD percent predicted 66.06 ± 12.45). The Kaplan-Meier analysis for group A patients is shown in Figure 1; 89% of patients with a normal baseline FVC did not decrease the percent-predicted FVC by ≥15 points at 5 years. In contrast, 75% of patients with a reduced baseline FVC did not decrease the percent-predicted FVC by ≥15 points at 5 years (P = 0.04). Four patients with decreased FVC at baseline developed severe respiratory failure (FVC <50% predicted) compared with none in the group with normal FVC at baseline (P < 0.05). Baseline FVC was also associated with subsequent DLCO impairment. The median time to DLCO decline to <70% of that predicted (Figure 2) was 6 years for patients with normal FVC at baseline and 2 years for patients with decreased baseline FVC (P = 0.001). A similar difference between the 2 subgroups was also found when severe DLCO impairment (<40% of that predicted) was used as the outcome (P = 0.01) (Figure 3). In this analysis, the DLCO of 89% of patients with a normal baseline FVC did not decrease <40% at 5 years from disease onset, whereas the DLCO of 68% of patients with reduced baseline FVC did not decrease <40% during the same time interval.

thumbnail image

Figure 1. Kaplan-Meier analysis of influence of initial forced vital capacity (FVC) measured within the first 3 years from disease onset on probability of subsequent decline of FVC in patients with scleroderma lung. Solid line indicates baseline FVC ≥80%; broken line indicates baseline FVC <80%.

Download figure to PowerPoint

thumbnail image

Figure 2. Kaplan-Meier analysis of influence of initial forced vital capacity (FVC) measured within the first 3 years from disease onset on probability of subsequent diffusing capacity for carbon monoxide (DLCO) decline (<70% of predicted) in patients with scleroderma lung. Solid line indicates baseline FVC ≥80%; broken line indicates baseline FVC <80%.

Download figure to PowerPoint

thumbnail image

Figure 3. Kaplan-Meier analysis of influence of initial forced vital capacity (FVC) measured within the first 3 years from disease onset on probability of subsequent severe diffusing capacity for carbon monoxide (DLCO) decline (<40% of predicted) in patients with scleroderma lung. Solid line indicates baseline FVC ≥80%; broken line indicates baseline FVC <80%.

Download figure to PowerPoint

The Kaplan-Meier analysis for group B patients whose pulmonary function measurements during the fourth or fifth year from disease onset were considered as baseline values is shown in Figure 4. In this group, 8 of 16 patients had normal FVC (mean ± SD percent predicted 89.6 ± 17) at baseline, and the remaining 8 patients had reduced FVC (mean ± SD 63.2 ± 11.4). As shown in Figure 4, these 2 subgroups did not differ in the probability of further decreasing FVC over 5 years (P = 0.13). Analysis of the remaining variables demonstrated that baseline DLCO, patient age, male sex, anti–topo I antibodies, and duration of Raynaud's phenomenon preceding skin manifestations were not found to be associated with pulmonary function loss in our patients (log rank P < 0.05 for all).

thumbnail image

Figure 4. Kaplan-Meier analysis of influence of initial forced vital capacity (FVC) measured after the first 3 years from disease onset on probability of subsequent decline of FVC in patients with scleroderma lung. Solid line indicates baseline FVC ≥80%; broken line indicates baseline FVC <80%.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We found that baseline FVC values measured within the first 3 years from disease onset may predict the subsequent rate of change in pulmonary function. Patients with a normal FVC early in the course of scleroderma have a greater likelihood of maintaining their pulmonary function than patients with an abnormal FVC at disease onset. In contrast, FVC values obtained late (>3 years) in the disease course have no predictive value.

In the present study, we used strict criteria to define pulmonary function loss. The threshold of an FVC decrement of at least 15 points (percent predicted) is more stringent than that used by other investigators (25) to characterize pulmonary function impairment, and is higher than the within-individual variation in pulmonary function results (25). Furthermore, because pulmonary function had to be sustained during all subsequent measurements, possible errors introduced by occasional improvements in pulmonary function described in certain patients with scleroderma without specific treatment (14, 26, 27) as well as errors related to the regression toward the mean were very unlikely.

We made a special effort to include a uniform patient group. First, all patients were white, therefore race-related factors that could potentially influence the rate of pulmonary function decline were lacking (2, 15, 17, 20). In addition, patients with clinical evidence of isolated pulmonary hypertension were not included in the analysis. The evolution of pulmonary function in patients with pulmonary hypertension differs significantly from that of patients with predominantly parenchymal fibrosis (28, 29). Finally, smoking status, which was found to be an important determinant of pulmonary function impairment in previous similar studies (13, 16, 17, 23), did not bias our data. There were 9 smokers, but all discontinued smoking around the time of disease onset and were equally distributed between the 2 groups.

Our study paid particular attention to the effects of timing of pulmonary function measurement on the overall ability of FVC to predict subsequent loss in lung function. The data clearly demonstrate that pulmonary function parameters obtained after a period of ∼3 years from disease onset have no predictive value. An obvious interpretation of this finding is most likely the difference in the annual rate of pulmonary function loss during the early course of scleroderma. In a study of a subset of patients with scleroderma with severe pulmonary restriction, Steen and colleagues (15) reported that the rate of FVC loss was ∼32% per year during the first 2 years of scleroderma and decreased to ∼3% annually after the initial 3 years. Our observations therefore can be easily interpreted if we consider the curve of pulmonary function decline as being initially steep and then relatively flat. Measurement of pulmonary function during the steep portion of the curve would predict a deteriorating function and development of respiratory failure. In contrast, if the initial measurement is performed during the flat portion of the curve, no matter how low pulmonary function is, it would not be associated with worsening pulmonary function and therefore would not be a predictor of pulmonary function loss.

Differences in the timing of pulmonary function measurement among studies may also explain previous conflicting data on the ability of initial pulmonary function measurement to predict further function loss. For example, in a prospective study in which the first pulmonary function assessment was made <3 years from disease onset, Morgan et al (6) demonstrated that pulmonary function at baseline was an important predictor of end-stage lung disease, and that a normal baseline function was associated with very low subsequent risk for developing respiratory failure. In contrast, in Greenwald et al's study (16) in which the interval of first pulmonary function measurement from disease onset was relatively long (average >10 years), the presence or degree of lung function impairment on initial testing was not associated with annual change of lung function. In agreement with our results, Greenwald et al (16) noted that long disease duration prior to study entry resulted in a slower annual decrease in lung volumes, and that such lung function data had no significant influence on the subsequent rate of change in lung function.

Baseline DLCO, male sex, patient age, and anti–topo I were not found to correlate with pulmonary function loss in the present study. In general, the ability of these parameters to predict function loss in patients with scleroderma is controversial. Initial values of DLCO did not predict progression to severe restrictive disease in the studies by Morgan et al (6) and Steen et al (14). Male sex was found to have a poor prognostic significance of pulmonary function loss in some studies (2, 18, 19) but not all studies (6, 16, 17). Although presence of anti–topo I is generally associated with lung involvement (1, 20, 21, 30), it has not been reported to correlate with greater pulmonary function loss over time in these patients. In a study involving >500 patients with scleroderma, Morgan et al (6) demonstrated that anti–topo I antibodies were not associated with greater function loss in these patients.

The limitations of the current study should be acknowledged. This was not a prospective study, and a greater proportion of patients with severe lung involvement and frequent pulmonary function testing may have skewed the results of the population studied. The results, therefore, need to be confirmed by prospective studies, which should include a large number of patients with scleroderma with various degrees of pulmonary involvement.

In conclusion, we found that FVC (percent predicted) measured within the first 3 years from disease onset is an independent predictor of pulmonary function deterioration in patients with scleroderma. Patients with an FVC ≥80% of that predicted have a greater likelihood of maintaining pulmonary function than patients with an FVC <80% of that predicted.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We thank Dr. E. Zintzaras for assisting with the statistical analysis.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES