Early Treatment of Stage II Sarcoidosis Improves 5-Year Pulmonary Function^*

Materials and Methods

Study Design

The 18-month treatment phase was a randomized, placebo-controlled, double-blind, parallel-group, multicenter study as previously described.⁵ The 5-year follow-up study was an open-label study extension.

Patients

The 18-month treatment phase included 189 randomized patients of both genders and ≥ 18 years old. They had newly detected (diagnosed within 3 months from the first symptoms or signs) radiographic stage I disease (bilateral hilar lymphadenopathy; n = 94) or stage II disease (parenchymal infiltrates with hilar lymphadenopathy; n = 84). Eleven patients were later reclassified as having had stage III disease (parenchymal infiltrates without hilar lymphadenopathy; n = 11) at the time of randomization. At baseline, normal lung function was seen in the majority of the patients. Only 23 patients had a FVC < 80% of predicted and 12 patients had a Dlco < 75% of predicted normal values. Ninety-two patients had been randomized to treatment with prednisolone followed by budesonide, and 97 patients were randomized to receive placebo. As previously reported, 154 patients completed the randomized, double-blind treatment phase⁵ and could be included in the follow-up study. The flowchart of patients with data from randomization to the end of the 5-year follow-up period is shown in Figure 1 separately for steroid-treated and placebo-treated stage I and stage II patients (the latter including the reclassified stage III patients).

Protocol

The investigational treatment was discontinued at 18 months. Thereafter, the patients visited the clinic once yearly (24 months, 36 months, 48 months, and 60 months after randomization) and underwent chest radiography, spirometry (FEV₁, FVC), Dlco, and biochemical testing (SACE, serum and urinary calcium). The detailed methods for all investigational procedures have been previously described.⁵ The interpretation of all chest radiographs was performed at the end of the 5-year follow-up by one of the investigators (A.P.) without actual information about previous treatment codes.

In the case of deterioration or relapse, individually adjusted treatments could be started during the follow-up. The type and length of treatments were recorded. At the clinic visits, patients were asked questions about adverse events.

Ethical Requirements and Good Clinical Practice

Ethics committees at each study center approved the study, including the 5-year follow-up, before any study-related procedures were undertaken. Clinical research nurses of Meltola Hospital and one of the investigators (A.P.) monitored the follow-up study.

Data Management and Statistical Analysis

All data were entered at Meltola Hospital, the coordinating center, into a New Mathematical Statistics Package 5.0x database. The 5-year data were analyzed using New Mathematical Statistics Package 5.05 (Firma NMSP Statistik program; Lund, Sweden). The results were analyzed based on an all-patient-treated approach. A “last value extended” principle was used when data from the 24-month or later visits were available but not from the last visit.

For statistical comparisons, t tests and analysis of variance were used. Separate analyses were performed to investigate the influence of the initial stages (I or II) on the outcome. The radiographic stages were analyzed by a stratified Wilcoxon test with strata according to stage at visit 1 (randomization visit). The stage was also analyzed by two-way analysis of variance with treatment and center as well as their interactions as factors. In all presentations of statistical tests results, a significant result refers to p values ≤ 5% (p < 0.05).

Results

Of the initial 189 randomized patients, 35 patients (16 steroid-treated and 19 placebo-treated) did not complete the 18-month treatment phase as previously reported.⁵ Of the 154 remaining patients, 149 patients participated in the follow-up part of the study. Five patients refused to take part in the follow-up. Figure 1 shows the initial randomization to treatment with corticosteroids or placebo, separately for patients with stage I and stage II(-III) pulmonary radiographic findings. The number of patients at the 5-year follow-up (n = 132) represents the total number of patients with complete data from randomization to the end of the follow-up period. However, all results are given for the 149 patients, including 70 corticosteroid-treated patients (31 women, 39 men) with a mean age of 39 years (range, 22 to 64 years) and 79 placebo-treated patients (38 women, 41 men) with a mean age of 41 years (range, 21 to 71 years). Randomization resulted in groups without statistically significant differences between them for age, gender, and initial chest radiographic stages.

Radiographic Findings

Table 1 shows the distribution of chest radiographic findings after the initial 18-month treatment phase and 5 years after randomization in 79 patients with initial stage I and 70 patients with initial stage II(-III) disease. In stage I disease, 31 of 39 patients in the steroid-treated group and 27 of 40 patients in the placebo group showed normal chest radiographic findings. Parenchymal lesions developed in six patients in each group. No statistically significant difference was found between the groups.

The corresponding developments in the 70 patients with initial stage II disease are also shown in Table 1. The nine stage III patients—five patients in the active-treatment and four patients in the placebo group—have been included in this evaluation. Twenty-one of 31 steroid-treated patients and 22 of 39 placebo-treated patients had normal chest radiographic findings. Ten steroid-treated patients and 17 placebo-treated patients had abnormal chest radiographic findings after 5 years of follow-up. However, no statistically significant differences in the development of chest radiographic findings between the groups could be found.

Lung Function Tests

The baseline values used in the following comparisons are the values of those patients followed up for 5 years. The mean baseline values therefore differ slightly from the mean values given in the previous publication.⁵

For stage I disease patients, no statistically significant differences were seen for FVC or Dlco between steroid-treated and placebo-treated patients (Table 2 ). For patients with initial stage II(-III) disease, the treatment effects were different (Table 3 ). In the corticosteroid-treated group, a mean increase in FVC of 0.33 L was observed at 18 months and this level was almost maintained at 5 years. In the placebo group, no change in FVC was seen over time. At both time points (18 months, 5 years), the difference in change in FVC between corticosteroid-treated and placebo-treated patients was statistically significant (p = 0.018 and p = 0.024, respectively). For Dlco, a mean increase was observed over time in the corticosteroid-treated group, whereas a mean decrease was seen in the placebo group. Also this difference in change over time was statistically significant (18 months, p = 0.034; 5 years, p = 0.028).

In the subgroups of patients with impaired lung function at baseline, the differences in changes in FVC and Dlco between corticosteroid-treated and placebo-treated patients were greater than in the groups of patients with a normal lung function. The differences between these groups did not, however, reach statistically significant levels because of the low numbers in the groups of patients with impaired lung function.

On the whole, the mean increases in FVC and Dlco in the corticosteroid-treated group are modest. The mean results are, however, a consequence of the selection procedure, with most included patients having a completely normal lung function at randomization. For individual patients with an initially impaired lung function, the improvements in lung function tests are considered clinically important.

Laboratory Test Values

The mean SACE and serum calcium levels as well as urinary calcium values obtained at baseline, after 18 months of treatment, and after 5 years of follow-up are shown in Tables 2,, 3 for initial stage I and stage II(-III) patients, respectively. No statistically significant differences were found between the groups.

Discontinuations and Treatments Administered During the Follow-up

After the 18-month treatment period, five patients never entered the follow-up phase. During the follow-up phase, corticosteroids had to be administered to 16 patients initially treated with placebo. In two patients initially treated with steroids, treatment had to be reinstituted because of a clinical relapse (Fig 2 ). The difference in frequencies between the groups in starting steroid treatment during the follow-up period was statistically significant (χ² = 12.17; p < 0.001).

The flow of patients during the entire 5-year study period is also illustrated in Figure 2. During the 18-month treatment phase, 19 patients discontinued the study as previously reported⁵ and 16 treatment failures were registered (6 in the active treatment group and 10 in the placebo group). During the subsequent follow-up period, 9 patients refused further follow-up visits and 18 patients required treatment with corticosteroids: 2 patients in the initial corticosteroid group because of relapses, and 16 patients initially treated with placebo. When the total number of treatment failures during the entire study is considered, the difference in outcome between the initial corticosteroid-treatment and placebo-treatment groups is also statistically significant (χ² = 10.49; p < 0.001).

Adverse Events

No serious adverse events were reported during the follow-up, and there were no discontinuations of the follow-up because of adverse events.

Discussion

Corticosteroids are considered to be beneficial in the treatment of pulmonary sarcoidosis, as their administration usually results in relief of respiratory symptoms and improvements in lung function and chest radiographic findings.⁸ Results of controlled clinical studies also show that both systemic⁹ and inhaled¹⁰¹¹ corticosteroid treatment influence cellular and biochemical findings, which appear to be important in the immunopathogenesis of the disease. However, reappearance of symptoms and radiographic infiltrates after discontinuation of treatment are frequent. As many as one third of treated patients have had recurrences within 2 years.¹²¹³ It is also noteworthy that results of optimally designed dose-response studies are lacking, and little is known about the necessary duration of treatment.

Although initial improvements are seen, many controlled clinical studies have failed to demonstrate a persisting difference in lung function tests and radiographic appearance between corticosteroid-treated and untreated patients.¹⁴¹⁵ The opinion has therefore often been expressed that corticosteroids do not alter the course of the disease and that treatment should be used only for relief of symptoms.² However, the opposite opinion also exists: corticosteroids should be administered with the aim of preventing irreversible fibrosis to patients with pulmonary infiltrates and active disease regardless of the presence or absence of symptoms.¹⁶¹⁷ Indeed, a number of clinical trials^141718192021 have clearly shown a beneficial effect of corticosteroid treatment compared with untreated or placebo-treated patients. The reason for the discrepancy in results between studies may be differences in patient populations, ethnic backgrounds, duration of disease with differences in the rates of spontaneous recoveries, and differences in study protocols, doses, and duration of treatment.

There are data to suggest that corticosteroids can impact on mechanisms important in the early stage of the pathogenic process. A number of genes, which play a role in the pathogenesis of sarcoidosis, can be modulated by corticosteroids. These genes are expressed soon after antigen triggering of CD4+ T cells and they include interleukin (IL)-2, interferon-γ, IL-1, and the IL-2 receptor, IL-2R.²² Other genes play important roles in the pathogenesis by influencing the expression of cell adhesion molecules (tumor necrosis factor-α) and leukocyte/monocyte proliferation and activation (granulocyte macrophage-colony stimulating factor, granulocyte-colony stimulating factor). They are also regulated by corticosteroids.^22232425

It could be postulated that early intervention is of importance when treatment is administered in order to suppress an inflammatory and immunologically mediated processes. Therefore, the aim of this study was to include only patients with newly detected pulmonary sarcoidosis, ie, with symptoms or signs of the disease for < 3 months. Patients fulfilling this criterion and with radiographic stage I and stage II lesions were included irrespective of lung function status and presence or absence of symptoms. The results of the 18-month treatment period showed no differences in stage I patients in radiographic development, lung function test results, or levels of biochemical serum markers.⁵ The frequency of spontaneous improvements in the placebo group was also higher than estimated in the power calculation of the study. However, in patients with initial stage II disease (11 patients were later reclassified as corresponding to stage III), a significant effect was seen in Dlco values and SACE activities.

This 5-year follow-up part of the study demonstrated in patients with initial stage I disease no significant differences between treated and untreated groups in any of the outcome variables. The vast majority of the patients had normal chest radiographic findings and normal lung function. This result confirms previous observations that stage I patients should be followed up without treatment until spontaneous recoveries can be documented. In patients with initial stage II disease, the result was different. Although no significant differences in chest radiographic developments could be detected, the differences in lung function tests were statistically significant. These differences also appear to be clinically relevant, indicating the value of early intervention with corticosteroids. There was no difference in the numbers of patients discontinuing the study during the initial treatment phase in the steroid-treated and the placebo-treated groups, respectively. During the follow-up phase, significantly more patients in the placebo group had to be treated, compared to the patients in the group initially randomized to treatment with prednisolone followed by budesonide. However, these patients treated during the follow-up phase were not excluded from the analysis. Therefore, any bias because of treatment during the study was in favor of placebo-treated patients and thus reducing the observed differences in lung function tests between the groups. However, it should be noted, that all functional abnormalities were small—and even so in patients with stage II disease—may be because of the short duration of disease (< 3 months) at the start of the study. Consequently, the treatment effects appear to be modest, too, although statistically significant.

It can be argued that treatment with corticosteroid may result in systemic steroid side effects. As many patients will recover spontaneously, an observation period of 6 to 12 months has been advocated.¹ The risk with this clinical approach is obviously that the subgroup of patients who require treatment will be treated too late. The treatment regimen used in this trial is therefore important, as the long-term maintenance treatment was inhaled medication with budesonide. The pharmacokinetic data speak in favor of a lung uptake high enough to result in a clinically meaningful anti-inflammatory effect. The metabolism of budesonide results rapidly in inactive metabolites; therefore, the risk of systemic side effects can be virtually eliminated. This means that no risk is taken even if some patients would be treated unnecessarily. Despite a favorable benefit-risk ratio, the cost-effectiveness of such a treatment approach needs further evaluations.

This 5-year follow-up study is the first to clearly demonstrate that early treatment of patients with pulmonary stage II sarcoidosis results in an improved functional outcome as measured by lung volumes and Dlco. Patients with stage I disease should be followed up without treatment.

Appendix

In addition to the authors, the following investigators and hospitals were members of the Finnish Study Group on Pulmonary Sarcoidosis, and participated in the study: Etelä-Karjala Central Hospital (R. Kauppinen), Etelä-Pohjanmaa Central Hospital (M. Koskenkari, E. Kokko, L. Tuomisto), Härmä Hospital (E. Aalto), Kanta-Häme Central Hospital (M. Järvinen), Keski-Pohjanmaa Central Hospital (R. Lillqvist, J-H. Slotte), Kiljava Hospital (R. Tammivaara, S. Koskinen), Kuopio University Hospital and Tarina Hospital (R. Majander), Kymenlaakso Central Hospital (M. Havu), Laakso Hospital (A. Ahonen, P. Saarelainen), Lappi Central Hospital (O. Säynäjäkangas), Länsi-Pohja Central Hospital (J. Kotaniemi), Oulu University Central Hospital (T. Keistinen, O. Rossi), Pietarsaari Hospital, Östanlid Department (J. Jaakkola), Pohjois-Karjala Central Hospital (E. Perämäki), Satalinna Hospital (V. Tuominen, S. Paloheimo, T. Saaresranta), Savonlinna Central Hospital (K.Venho), Tampere University Central Hospital and Pikonlinna Hospital (M. Nieminen), Turku University Central Hospital and Paimio Hospital (A. Palojoki, P. Laurinen), Vaasa Central Hospital (T. Rinne, L. Lammi).