Professor Denton has served on the advisory board of and received speaking fees (less than $10,000) from Actelion Pharmaceuticals. Professor Black has received consulting fees (less than $10,000) from Actelion Pharmaceuticals. Professor du Bois has served on the steering committee of and received speaking fees (less than $10,000) from Actelion Pharmaceuticals.
A multicenter, prospective, randomized, double-blind, placebo-controlled trial of corticosteroids and intravenous cyclophosphamide followed by oral azathioprine for the treatment of pulmonary fibrosis in scleroderma
Version of Record online: 28 NOV 2006
Copyright © 2006 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 54, Issue 12, pages 3962–3970, December 2006
How to Cite
Hoyles, R. K., Ellis, R. W., Wellsbury, J., Lees, B., Newlands, P., Goh, N. S. L., Roberts, C., Desai, S., Herrick, A. L., McHugh, N. J., Foley, N. M., Pearson, S. B., Emery, P., Veale, D. J., Denton, C. P., Wells, A. U., Black, C. M. and du Bois, R. M. (2006), A multicenter, prospective, randomized, double-blind, placebo-controlled trial of corticosteroids and intravenous cyclophosphamide followed by oral azathioprine for the treatment of pulmonary fibrosis in scleroderma. Arthritis & Rheumatism, 54: 3962–3970. doi: 10.1002/art.22204
- Issue online: 28 NOV 2006
- Version of Record online: 28 NOV 2006
- Manuscript Accepted: 9 AUG 2006
- Manuscript Received: 8 FEB 2006
- Arthritis Research Campaign. Grant Number: 14791
- Raynaud's & Scleroderma Association, UK
The lack of randomized controlled trials (RCTs) in pulmonary fibrosis in systemic sclerosis (SSc) has hampered an evidence-based approach to treatment. This RCT was undertaken to investigate the effects of intravenous (IV) cyclophosphamide (CYC) followed by azathioprine (AZA) treatment in pulmonary fibrosis in SSc.
Forty-five patients were randomized to receive low-dose prednisolone and 6 infusions (monthly) of CYC followed by oral AZA, or placebo. Primary outcome measures were change in percent predicted forced vital capacity (FVC) and change in single-breath diffusing capacity for carbon monoxide (DLCO). Secondary outcome measures included changes in appearance on high-resolution computed tomography and dyspnea scores. An intent-to-treat statistical analysis was performed.
At baseline, there were no significant group differences in factors linked to outcome, including severity of pulmonary fibrosis and autoantibody status. Sixty-two percent of the patients completed the first year of treatment. Withdrawals included 9 patients (6 from the placebo group) with significant decline in lung function, 2 with treatment side effects (both from the active treatment group), and 6 with non–trial-related comorbidity. No hemorrhagic cystitis or bone marrow suppression was observed. Estimation of the relative treatment effect (active treatment versus placebo) adjusted for baseline FVC and treatment center revealed a favorable outcome for FVC of 4.19%; this between-group difference showed a trend toward statistical significance (P = 0.08). No improvements in DLCO or secondary outcome measures were identified.
This trial did not demonstrate significant improvement in the primary or secondary end points in the active treatment group versus the group receiving placebo. However, for FVC there was a trend toward statistical significance between the 2 groups. This suggests that treatment of pulmonary fibrosis in SSc with low-dose prednisolone and IV CYC followed by AZA stabilizes lung function in a subset of patients with the disease. Therapy was well tolerated with no increase in serious adverse events.
Pulmonary disease in systemic sclerosis (SSc; scleroderma) is associated with significant morbidity and mortality (1). At least one-third of patients with SSc have clinically significant pulmonary fibrosis, and lung function impairment is evident in up to 70% (2). The 10-year survival rate from the time of presentation with pulmonary fibrosis in SSc approximates 70%, and many patients experience disabling progressive breathlessness (3). Although the pathogenesis of SSc-associated pulmonary fibrosis is incompletely defined, it is believed that chronic inflammation leads to progressive lung injury and incremental fibrosis (4–8). To date, treatments targeting the inflammatory pathway have included corticosteroids and immunosuppressive agents. This therapeutic approach has been largely empirical, paralleling treatment strategies used in idiopathic pulmonary fibrosis (IPF) (9); there is a paucity of randomized controlled trials (RCTs) in both diseases.
Cyclophosphamide (CYC) is the most studied immunosuppressive agent in pulmonary fibrosis in SSc. In previous uncontrolled studies, CYC has variably improved lung function parameters, including forced vital capacity (FVC) levels (6, 10–20). In a recent retrospective case series, azathioprine (AZA) therapy was associated with marginal reductions in dyspnea and stabilization of lung function parameters, especially FVC (21). These suggestive but inconclusive data were believed to justify a placebo-controlled study. We therefore designed a multicenter, prospective, randomized, double-blind, placebo-controlled trial of low-dose prednisolone and intravenous (IV) CYC, followed by oral AZA, in SSc-associated pulmonary fibrosis. This choice of regimen was influenced by previous uncontrolled trials in pulmonary fibrosis in SSc, our current practice at the Royal Brompton Hospital Interstitial Lung Disease Unit, and current international practice in IPF, in which AZA has a role in maintenance therapy (22, 23). IV CYC was chosen in an attempt to induce remission using a route of delivery that we hoped would reduce the side effect frequency. In this report we present the findings of the 1-year data analysis.
PATIENTS AND METHODS
Between January 1999 and March 2003, adult patients were recruited from the following 5 centers in the UK: Royal Brompton Hospital, Royal Free Hospital, Leeds General Infirmary, Royal National Hospital for Rheumatic Diseases/Royal United Hospital, and Hope Hospital. SSc patients with early pulmonary fibrosis, including new referrals and patients already being cared for at the participating study centers, were recruited.
To be included in the study, patients had to be age 18–75 years, fulfill the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) preliminary criteria for a diagnosis of SSc (24), have SSc-associated pulmonary fibrosis, as indicated by high-resolution computed tomography (HRCT) or thoracoscopic lung biopsy, and comply with therapy and with regular specialty center attendance. Patients were excluded from the study if they had had previous AZA or CYC therapy for >3 months, had had previous high-dose oral corticosteroid therapy (30 mg of prednisolone or equivalent daily) for >3 months, had had oral corticosteroid therapy (prednisolone dosage ≥10 mg daily) in the 3 months before study entry, had contraindications to oral corticosteroids such as poorly controlled diabetes or severe osteoporosis, were likely to require lung transplantation within 1 year, had a history of or laboratory data suggestive of other serious systemic or psychological disease unrelated to SSc, were pregnant or lactating, exhibited evidence of alcohol or drug abuse, or were unable to give written informed consent.
Multicentre Research Ethics Committee approval (MREC/99/2/116) was obtained, and each participating center received local ethics committee approval. All patients provided written informed consent. The study was conducted in accordance with the Declaration of Helsinki (Hong Kong amendment, 1989).
Routine baseline investigations, performed prior to randomization, included chest radiography, HRCT, and pulmonary function tests. HRCT sections (1.5-mm cuts at 20-mm intervals from lung apex to base in the supine position) were graded using a semiquantitative scoring system. Two thoracic radiologists scored sections at 5 predetermined levels, estimating the total extent of disease (to the nearest 5%) and the proportion of ground-glass attenuation (25). The radiologists were blinded to treatment category and patient detail, and consensus scores were used if there was a discrepancy between scorers.
Pulmonary function testing was performed according to the guidelines of the British Thoracic Society (26), using Compact Masterlab equipment (Jaeger Healthcare, Hoechberg, Germany). Tests of lung volume and airway resistance were conducted using a constant volume body plethysmograph. Single-breath diffusing capacity for carbon monoxide (DLCO) was tested using a conventional carbon monoxide/helium gas mixture and corrected for hemoglobin (corrected DLCO) and alveolar volume (coefficient of gas transfer, Kco). Spirometry was performed using flow-volume loops. All results were expressed as a percentage of normal predicted values based on age, sex, and height. A modified American Thoracic Society respiratory questionnaire was used to determine a baseline dyspnea score; the minimum score of 0 represented breathlessness after 30 minutes of vigorous activity, and the maximum score of 20 represented breathlessness with minimal activity (27).
Other baseline variables measured included autoantibody status (the presence or absence of the antitopoisomerase antibody), resting arterial blood gases, electrocardiography findings, echocardiography findings, urinalysis results, and results of blood tests including determinations of urea, electrolyte, and serum glucose levels, liver function tests (LFTs), and complete blood cell count (CBC). Bronchoalveolar lavage and 99mTc–diethylenetriaminepentaacetic acid (DTPA) clearance testing in nonsmokers were performed according to local practice; a surgical biopsy was performed if clinical and HRCT features were not classical of pulmonary fibrosis in SSc.
Patients were allocated to either the active treatment group (prednisolone and CYC followed by AZA) or the placebo group, using the minimization method with balancing for the following known prognostic factors: age, baseline HRCT pattern and extent of disease, and autoantibody profile (28). Investigators were blinded to the treatment allocation. Randomization was undertaken at the Royal Brompton Hospital by members of the Clinical Trials and Evaluation Unit, who were not involved in the analysis of data.
Patients in the active treatment group received therapy with 20 mg oral prednisolone on alternate days and 6 IV infusions of CYC at a dose of 600 mg/m2 (mean dose 1,050 mg) at 4-week intervals, followed by oral AZA at 2.5 mg/kg/day (maximum 200 mg/day) as maintenance therapy. Patients in the placebo group received placebo formulations that matched the active treatment. High fluid intake and regular bladder voiding were recommended during the CYC/placebo infusion to minimize the risk of hemorrhagic cystitis; in our experience, we have not found it necessary to prescribe mesna for patients receiving CYC at these doses. AZA was prescribed initially at a test dosage of 50 mg/day, and was increased to the full therapeutic dosage after 1 month in the absence of significant adverse events. Both groups continued to receive standard therapy for nonpulmonary disease. Other immunosuppressive agents, including colchicine, methotrexate, and D-penicillamine, which might influence the course of the lung fibrosis, were not permitted.
Patients were reviewed monthly for the first 6 months, and at 9 and 12 months. Blood tests comprising CBC, urea, electrolytes, LFTs, and serum glucose (to exclude steroid-induced diabetes) were performed immediately prior to the placebo/CYC infusion for the first 6 months, and at each subsequent visit in both treatment groups. Urinalysis (Multistix SG; Bayer, Bury St. Edmunds, UK), followed by formal microscopy in the event of an abnormal result, was also performed immediately prior to the placebo/CYC infusion for the first 6 months, and at each subsequent visit in both treatment groups. Tests of lung function and arterial blood gases were performed at 3, 6, and 12 months; HRCT was repeated at 1 year.
In addition, during the initial period of treatment with AZA (or matched placebo), CBC and LFTs were performed weekly for 4 weeks, once every 2 weeks for 4 weeks, and every 6 weeks thereafter to monitor for potential bone marrow suppression or hepatotoxicity. Serial HRCT change was quantified by a modified progression scoring system originally described by Antoniou et al (29). Images were assessed by 2 observers (who were blinded to treatment category and sequence) over 2 anatomic zones; the upper region extended from the main carina to the section midway between the pulmonary venous confluence and the top of the right hemidiaphragm, and the lower region extended caudally to the costophrenic angles. The 7-point scale ranged from 1 (major improvement) to 7 (major deterioration). Improvement was defined as any score <4.
Efficacy end points.
The primary end points were change in percent predicted FVC and corrected DLCO. Secondary outcome measures were change in dyspnea scores (≥1 grade), sustained across 2 time points, and change in HRCT extent and pattern of disease at 1 year.
If lung function deterioration, defined as a 10% drop in absolute FVC or a 15% drop in absolute corrected DLCO, sustained over 2 time points ≥1 month apart, was observed during the treatment period, the study treatment was discontinued and, after unblinding, the best alternative treatment in the opinion of the local clinician was offered.
Power calculation was based on detecting a difference of ≥1 SD between the mean values in the 2 treatment groups. Such a difference can be detected at the 0.05 level of statistical significance with 90% power with a sample size of 22 patients per group. Outcome assessment at 1 year was based on percent predicted values for FVC, corrected DLCO, total lung capacity, forced expiratory volume in 1 second, and coefficient of gas transfer, adjusted for age, sex, and height. The treatment effect of the active therapy compared with control was estimated using analysis of covariance adjusted for the baseline value of the outcome measure and treatment center. Assumptions in the statistical analyses were checked using normal probability and leverage plots. Analysis was based on intent-to-treat subject to the availability of data at 1 year.
Forty-five patients (13 men, 32 women) ages 18–75 years (median age at randomization 55 years) were enrolled in the study; 22 patients (49%) were assigned active treatment with prednisolone and CYC followed by AZA, and 23 patients (51%) were assigned placebo. All patients met the ACR diagnostic criteria for SSc (24). The trial profile is outlined in Figure 1.
Table 1 summarizes the baseline characteristics of the patients by treatment group. The patient groups were similar in age, sex, ethnicity, and smoking status. Both patients with limited cutaneous SSc (lcSSc) and patients with diffuse cutaneous SSc (dcSSc) were included in the study. In the active treatment group, 15 patients had lcSSc and 7 had dcSSc; in the placebo group, 14 had lcSSc and 9 had dcSSc. The interval between a formal diagnosis of pulmonary fibrosis and randomization was a median of 4 months in both groups (range 1–55 months in the active treatment group, 1–59 months in the placebo group). The interval between the diagnosis of SSc, defined as the onset of the first non-Raynaud's symptom, and randomization was a median of 33 months in the active treatment group (range 1–204 months) and a median of 66 months in the placebo group (range 3–322 months).
|Characteristic||Active treatment group (n = 22)†||Placebo group (n = 23)|
|Royal Brompton/Royal Free||17 (77.3)||13 (56.6)|
|Bath||1 (4.5)||6 (26.0)|
|Salford||2 (9.1)||0 (0)|
|Leeds||2 (9.1)||4 (17.4)|
|Male||5 (22.7)||8 (34.8)|
|Female||17 (77.3)||15 (65.2)|
|White||20 (90.9)||21 (91.4)|
|Black||0 (0)||1 (4.3)|
|Asian||2 (9.1)||1 (4.3)|
|Current smoker||3 (13.6)||3 (13.0)|
|Ex-smoker||7 (31.8)||5 (21.8)|
|Nonsmoker||12 (54.6)||15 (65.2)|
|Days to outcome assessment, mean (range)‡||364.4 (305–439)||358.7 (223–414)|
The predominant scleroderma-specific autoantibody identified was antitopoisomerase antibody, found in 45.5% of the patients in the active treatment group and 34.8% of the patients in the placebo group; other serologic findings included positivity for anticentromere antibody (in 4.5% of patients in both groups), anti–RNA polymerase I, II, III antibody (in 8.7% of patients in the placebo group), anti-RNP antibody (in 4.5% of patients in both groups), anti–PM-Scl antibody (in 4.5% of patients in the active treatment group), perinuclear antineutrophil cytoplasmic antibody (in 4.5% of patients in the active treatment group), and anti–Jo-1 antibody (in 4.5% of patients in the placebo group). The remainder of the patients were found to have an uncharacterized antinuclear antibody. The mean number of days until followup assessment was similar in the 2 groups.
Baseline pulmonary evaluation.
At baseline, no significant differences between groups were identified in pulmonary disease severity as characterized by lung function tests, HRCT pattern and extent of disease, gas exchange values, baseline hemodynamics, bronchoalveolar lavage cellularity, and DTPA clearance (Tables 2 and 3). Both patients with lcSSc and patients with dcSSc had similar disease severity at baseline, and were equally matched between groups; overall, in the lcSSc subset, the mean percent FVC was 82% and the mean percent corrected DLCO was 55%, whereas in the dcSSc subset, the mean percent FVC was 78% and the mean percent corrected DLCO was 51%. Patients in both the active treatment and placebo groups who carried the antitopoisomerase antibody had a mean percent FVC of 78%, whereas those patients in both groups who were antitopoisomerase antibody negative had a mean percent FVC of 82%. The percent corrected DLCO was similarly matched between patients who had the antitopoisomerase antibody and patients who did not. In view of atypical clinical or HRCT features, surgical lung biopsies were performed in 5 patients (11%), 4 in the active treatment group and 1 in the placebo group; all biopsies revealed a histopathologic pattern of nonspecific interstitial pneumonia. Echocardiography, performed in 42 patients (93%), excluded pulmonary hypertension in all cases. Thus, no significant differences were identified between groups in a number of factors linked to outcome in pulmonary fibrosis in SSc.
|Treatment group||Placebo group|
|DTPA clearance, minutes†||29.4 (13.3–55.5) ||36.5 (14.0–115.0) |
|Neutrophils, % total cells||11.7 (2.0–45.5) ||13.9 (4.0–28.6) |
|Eosinophils, % total cells||6.9 (0.4–31.0) ||8.18 (0.0–24.3) |
|Lymphocytes, % total cells||11.1 (1.7–24.0) ||11.7 (2.6–30.3) |
|Arterial blood gases|
|PO2, KPA||11.22 (10.00–13.50) ||11.53 (8.76–13.43) |
|PCO2, KPA||5.49 (4.05–6.00) ||4.97 (4.23–6.61) |
|Antitopoisomerase antibody positivity, no. (%)||10 (45.5)||8 (34.8)|
|Treatment group (n = 22)||Placebo group (n = 23)||Treatment group (n = 19)||Placebo group (n = 18)||P†|
|Lung function, % predicted|
|FVC||80.1 ± 10.3||81.0 ± 18.8||82.5 ± 11.3||78.0 ± 21.6||0.08|
|DLCOc||52.9 ± 11.5||55.0 ± 12.9||49.6 ± 10.7||51.8 ± 14.9||0.64|
|TLC||81.8 ± 10.1||76.8 ± 16.9||80.2 ± 9.8||74.4 ± 16.7||0.61|
|FEV1||79.6 ± 11.5||79.7 ± 19.1||81.3 ± 12.5||77.0 ± 21.3||0.16|
|Kco||71.3 ± 13.4||82.7 ± 19.1||71.5 ± 13.9||77.9 ± 23.3||0.32|
|Disease extent, mean (range) %||20 (6–40)||19 (5–40)||–||–||–|
|Ground-glass attenuation, mean (range) %||50 (15–91)||47 (0–95)||–||–||–|
|Improvement on serial HRCT, no (%)‡||–||–||6 (40)||3 (20)||0.39|
|Dyspnea score, mean (range)§||7.7 (2–14)||7.2 (0–18)||8.75 (0–14)||7.80 (2–14)||0.23|
Twenty-eight of 45 patients (62%) completed the first year of the study, as illustrated in the trial profile (Figure 1). Seventeen of the 45 (38%) were withdrawn in the first year (7 patients in the active treatment group and 10 in the placebo group). Significantly, 9 patients (3 in the active treatment group, 6 in the placebo group) were withdrawn due to a significant decline in FVC or corrected DLCO (>10% or >15% of absolute value, respectively), whereas only 2 were withdrawn due to treatment side effects (both in the active treatment group). Other withdrawals were due to active extrapulmonary disease requiring immunosuppressive treatment, which violated the study protocol (3 patients, all in the placebo group), poor IV access (1 patient in the active treatment group), and thrombocytopenia secondary to a new diagnosis of follicular lymphoma (1 patient in the placebo group). One patient (in the active treatment group) was withdrawn after only 39 days in the trial due to investigation for metastatic malignancy with an unknown primary site; the short duration of active therapy (1 IV pulse CYC) makes a causal link with immunosuppression extremely unlikely.
Eight of 45 patients (18%) were lost to followup (3 patients in the active treatment group and 5 in the placebo group). All of these patients had previously been withdrawn from the study. For patients who withdrew from the study or for whom unblinding occurred, outcome data were obtained either at the withdrawal appointment or via subsequent contacts as close to the 1-year anniversary of randomization as possible. In 5 cases outcome data were obtained at 2 appointments before and after the 1-year time point, and were used to linearly interpolate a 1-year value. Therefore, 1-year data on 19 of 22 patients (86%) in the active treatment group and on 18 of 23 patients (78%) in the placebo group were available to be included in an intent-to-treat analysis.
Efficacy end points.
Table 3 summarizes the outcomes in each group in terms of efficacy end point variables. The difference in the change in FVC in the active treatment group versus the placebo group showed a trend toward statistical significance (P = 0.08). Estimation of the relative treatment effect (study treatment versus placebo) after adjustment for baseline FVC (i.e., baseline disease severity) and treatment center revealed a difference (improvement) of 4.19% in the predicted FVC in the active treatment group (95% confidence interval [95% CI] −0.57, 8.95; P = 0.08). This represents a mean unadjusted improvement of 2.4% in predicted FVC in the active treatment group, and a mean unadjusted decline of 3.0% in predicted FVC in the placebo group. No significant or marginal (0.05 < P < 0.10) improvements in other lung function parameters were identified in the active treatment group compared with the placebo group.
Since obesity, defined as weight >80 kg, was seen in ≥20% of the study population, a post hoc subanalysis with additional adjustment for weight was performed, in view of the possibility that obesity would impact the FVC due to increased thoracic load. In this subanalysis, the estimated treatment effect was a difference (improvement) of 4.76% in the predicted FVC in the active treatment group (95% CI −0.14, 9.38). This difference achieved statistical significance (P = 0.04). With regard to the natural history of untreated disease, similar rates of lung function decline were seen in both the limited and diffuse disease subsets in the placebo group, and, importantly, patients carrying the antitopoisomerase autoantibody did not have a greater rate of lung function decline, reducing the potential for bias.
With respect to the secondary outcome measure of HRCT pattern and extent of disease, 30 HRCTs were available to serially score at 1 year, 15 in each group (68% in the active treatment group and 65% in the placebo group). Six of 15 patients (40%) in the active treatment group demonstrated a degree of improvement (reduced coarseness and/or extent of disease), compared with only 3 of 15 patients (20%) in the placebo group; this trend toward improved HRCT appearance did not reach statistical significance (P = 0.39). No improvement with active therapy was identified using serial dyspnea scores. One death occurred during the first year; the patient in the active treatment group who was diagnosed as having metastatic malignancy on day 39, and subsequently withdrawn from the study, died after 5 months.
Adverse events with active treatment were few, generally not sustained, and resulted in withdrawal in only 2 instances; 1 patient developed intolerable nausea during treatment with CYC, whereas the other developed abnormal findings on LFTs during treatment with AZA. Transient adverse events included nausea in 8 of 22 patients (36.4%), mood disturbance in 4 patients (18.2%), mouth ulceration in 3 patients (13.6%), rash in 3 patients (13.6%), abnormal findings on LFTs in 2 patients (9.1%), diarrhea in 2 patients (9.1%), and dyspepsia in 1 patient (4.5%). Serious adverse events occurred at similar rates in both groups. Intercurrent respiratory tract infections occurred in 3 of 22 patients (13.6%) in the active treatment group and in 4 of 23 patients (17.4%) in the placebo group. One of the latter patients required hospital admission. One patient in each group developed a malignancy and was withdrawn, as detailed above. Bedside urinalysis (Multistix) testing suggested the presence of hematuria in 3 of 22 patients (13.6%) in the active treatment group and 4 of 23 patients (17.4%) in the placebo group at baseline, and in 10 of 22 patients (45.5%) in the active treatment group and 6 of 23 patients (26.1%) in the placebo group over the subsequent followup period. Reassuringly, hematuria (defined as >10 cells/mm3) was confirmed by formal urine microscopy in only 3 patients (1 in the active treatment group and 2 in the placebo group) and was persistent only in 1 (in the placebo group), where a prior diagnosis of SSc-related renal crisis and interstitial nephropathy had been made. No episodes of hemorrhagic cystitis or bone marrow suppression due to therapy with immunosuppressive agents were identified.
This RCT was designed to examine the efficacy of low-dose prednisolone and IV CYC followed by oral AZA in patients with pulmonary fibrosis and SSc; the lung function response to active therapy approached, but did not achieve, conventional statistical significance. The response of the primary end point, FVC, to active treatment (P = 0.08) was consistent with findings recently demonstrated in the Scleroderma Lung Study (SLS), in which FVC was also chosen as a primary outcome measure, and a treatment advantage with oral CYC was found (P = 0.03) (30). These results support the trends identified in previous uncontrolled, open studies with IV and oral CYC (6, 10–13, 17). Taken together, the findings of these 2 prospective studies (the present study and the SLS) suggest that CYC is beneficial for lung function over 1 year in patients with SSc.
The clinical significance of the relatively small changes in lung function found in both studies needs to be addressed. The long-term implications of small changes in FVC at 1 year are unknown, and it is not known whether these findings will translate into improved survival. In addition, it is not known whether this trend predicts subsequent change. However, we would suggest that a trend for improvement, intuitively better than a trend for decline, is hugely encouraging for the patient, and this must not be ignored. In addition, should this trend persist over a second and possibly further years, then it is likely to presage a more positive outcome based on historical data.
Our findings are consistent with a combination of treatment response and prevention of decline, and highlight the need to define outcome in fibrosing lung disease in terms of changes in trend, particularly with respect to lung function, when assessing therapeutic efficacy. The response to therapy in diseases in which fibrosis is a significant component is not characterized by large shifts in lung function indices, and the presence of a placebo group, absent in the vast majority of previous studies, has allowed us to demonstrate divergent trends between groups where the change is relatively small. The amplitude of effect demonstrated in this study is very similar to that found in the SLS; the discrepancy in statistical significance reflects our smaller sample size.
Inherent in the design of placebo-controlled trials in SSc-associated pulmonary fibrosis is the selection of patients with atypically nonprogressive and mild disease; trial enrollment in these cases is more acceptable to both patients and physicians. This may be particularly true in studies of CYC, a therapeutic agent that is widely advocated based on good anecdotal evidence. In addition, many patients with severe or deteriorating disease were inevitably excluded from this study, since a recent history of immunosuppressive therapy or moderate-dose prednisolone treatment would fulfill exclusion criteria.
If the major benefit of a therapy is prevention of disease progression, and selection factors cause patients with less severe and progressive disease to be studied, the therapeutic effect must necessarily be small and may seriously understate a potential benefit in more progressive disease. This unavoidable issue may have attenuated the significance of the primary end points. As an illustration, the mean baseline percent predicted FVC in the combined groups in this study was 81.0%, whereas the mean baseline percent predicted FVC in previous uncontrolled studies of oral CYC, in which significant improvements in FVC were demonstrated, was as low as 60% (10, 12–13). The population selected for the present study was characterized by mild functional impairment and, thus, only a minor therapeutic benefit was likely to be attainable. It appears plausible that the observed therapeutic benefit might translate to a clinically important effect in patients with more severe or progressive disease, although this study was not designed to address this group for the reasons stated above.
In <25% of patients, a significant decline in lung function parameters necessitated the open introduction of standard treatment, as rescue medication. Because followup data used in the intent-to-treat analysis included data obtained following rescue medication, the differential benefit of active therapy may have been understated.
The majority of our patient population was considered for inclusion in the trial when physiologic or radiologic lung abnormalities were first identified (i.e., relatively early pulmonary disease). Historically, there has been little evidence to support the use of immunosuppressive therapy at this stage of disease; this study highlights the potential advantages of initiating a therapy that has few side effects before patients become functionally limited, thus allowing them to retain a good quality of life.
Choice of the route of delivery of CYC is influenced by issues of toxicity; on balance, IV administration appeared to confer an advantage over the oral administration used in previous uncontrolled studies at our institution. Significantly less toxicity, particularly with regard to the eventual risk of bladder cancer, is believed to occur following intermittent IV treatment (14–18). IV CYC has become acceptable therapy for other systemic rheumatic diseases and for Wegener's granulomatosis, with less morbidity being observed than with the oral regimen (31). Therefore, in view of the potentially favorable therapeutic-to-toxicity ratio with this route of delivery, and in an effort to minimize variable absorption, we used IV CYC in this study.
AZA, which has been shown to improve survival in IPF and is also associated with a favorable side effect profile, was used as maintenance therapy, in order to avoid the need for ongoing parenteral treatment. The concurrent use of corticosteroids reflects current consensus management in both IPF and SSc-associated pulmonary fibrosis, although RCT data are lacking; the steroids were administered in low doses to avoid the potentially increased risk of scleroderma renal crisis (32). Therapy was well tolerated, with transient nausea, rashes, hepatic dysfunction, and diarrhea in a very small proportion of patients. Only 2 patients withdrew from active therapy due to adverse events, and, reassuringly, no immunosuppressive agent–induced bone marrow suppression or hemorrhagic cystitis was identified.
Longitudinal observational data on both IPF and SSc-associated pulmonary fibrosis has increased in recent years, and serial FVC has come to be recognized as the most accurate and reproducible monitoring tool (33–35). The role of serial DLCO in monitoring fibrosing lung diseases has been widely debated. Once seen as the gold standard in monitoring for disease decline, its value is hampered by difficulties with standardization and by concurrent, often subclinical, pulmonary vascular disease in SSc patients with pulmonary fibrosis. In this study, a significant change in DLCO was not observed, consistent with the findings of the majority of previous uncontrolled studies (6, 10–12, 14–15).
No significant difference between groups was found for the secondary end points, dyspnea score and HRCT appearance. Patients with pulmonary fibrosis in SSc often have extrathoracic musculoskeletal involvement, which may increase their respiratory effort on exertion. It is possible that this may have had an impact on dyspnea scores; we have certainly observed this in our patients. With respect to the absence of significant change in HRCT appearance, it has been increasingly recognized that ground-glass attenuation, particularly in the presence of traction bronchiectasis, may reflect fine fibrosis, rather than simply inflammation or alveolitis as once presumed (36). In this regard, gross changes in HRCT appearance may be too nonspecific as a marker of treatment response.
Statistical analyses were performed on an intent-to-treat basis, and thus potential bias was minimized. Specifically, every effort was made to obtain 1-year data on all patients, and data were obtained for all but 8, who were lost to followup. Blinding was maintained throughout the data analysis. Maintenance of the blinding was facilitated by the relatively infrequent occurrence of adverse events, in contrast with the frequency of adverse events in earlier studies of oral CYC (10, 13).
In summary, we have described an RCT examining the effect of low-dose prednisolone and IV CYC followed by oral AZA in pulmonary fibrosis in SSc. The treatment effect approached, but did not achieve, conventional statistical significance, and the treatment was associated with good drug tolerance and compliance. Although these findings are consistent with recent findings demonstrated in the SLS, a placebo-controlled study of oral CYC (30), the present study has some limitations. First, the study population was a subset of patients with pulmonary fibrosis in SSc whose disease was in a relatively stable phase, and IV delivery of CYC was not directly compared with oral delivery. Second, the sample size was limited, which may in part explain the failure to achieve statistical significance. In isolation, our study does not provide unequivocal evidence that this regimen is the ideal treatment plan at this time; however, it is one option for treatment. For the first time, prospective, randomized, placebo-controlled studies have helped to clarify the role of immunosuppression in the treatment of diffuse lung disease occurring in the context of a connective tissue disease.
We are grateful to trial coordinators K. Griffin (Hope Hospital), S. Butts (Royal United Hospital for Rheumatic Diseases and Royal United Hospital), N. Waldron (Royal United Hospital for Rheumatic Diseases and Royal United Hospital), and N. Reay (Leeds General Infirmary).
- 27American Thoracic Society. Surveillance for respiratory hazards in the occupational setting. Am Rev Respir Dis 1982; 126: 952–6.