Update in systemic sclerosis
Systemic sclerosis (SSc; scleroderma), is a multiorgan systemic disease. Patients develop organ dysfunction secondary to a mixture of fibrotic and vascular disease in the setting of underlying autoimmunity. Although cutaneous fibrosis is the disease hallmark, fibrosis of visceral organs (such as the lungs), vascular damage, and intimal vascular occlusion and fibrosis are the leading causes of morbidity and mortality.
The articles included in this review are recently published studies (1999–2002) in SSc that are relevant in teaching us about the pathophysiology of the disease, natural history of certain target organ damage, new therapies that are currently available, and therapies on the horizon.
Renal functional reserve is impaired in patients with systemic sclerosis without clinical signs of kidney involvement (Ann Rheum Dis, 2002) (1)
The hallmark of renal involvement in SSc is scleroderma renal crisis, the abrupt onset of renal failure usually with associated severe hypertension. The mortality rate in SSc has improved significantly over the years due to improvements in management of severe renal disease. This study was undertaken to evaluate renal function in SSc patients without renal crisis. The functional response of the kidney to an amino acid challenge (called the renal functional reserve [RFR]) was measured in asymptomatic nonhypertensive scleroderma patients versus a control population.
Twenty-one patients were selected from all patients with scleroderma who were referred to the authors' medical center over a 6-month period. Inclusion criteria included diastolic blood pressure ≤90 mmHg at multiple measurements, serum creatinine <92 μmol/liter, 24-hour creatinine clearance >70 ml/minute × 1.73 m2, no clinical proteinuria, and normal urinary sediment. No patient had been treated with an angiotensin converting enzyme (ACE) inhibitor or an angiotensin II receptor blocker. Ten healthy controls were included in the study. Of the 21 patients with scleroderma, 5 had diffuse disease and 16 had limited disease. The interval between the first diagnosis of SSc and entry into the study ranged from 1 to 76 months (mean 28.1, median 24 months).
After baseline evaluation of creatinine clearance and protein excretion, patients were subjected to amino acid challenge. After an overnight fast, with patients in the upright position, renal function was measured in the baseline condition and after administration of an amino acid load (Freeamine III; Baxter Diagnostic, McGaw Park, IL; 8.5% solution, 4.16 ml/minute for 2 hours). Serum concentrations of sodium, potassium, urea, and creatinine; serum osmolality; cell volume; and plasma renin activity were measured. Creatinine and PAH clearance were used to estimate the glomerular filtration rate (GFR) and effective renal plasma flow (ERPF).
In the control group, the amino acid load induced a marked increase in creatinine clearance (+29.0 ml/minute) and sodium paraaminohyppurate PAH clearance (+111.7 ml/minute), which reflects an increase in GFR (of +30.0 ml/minute [SD 13.3]). The total renal vascular resistance (TRVR) also decreased significantly (−1,659 dynes/second × cm5) in the control group. In the scleroderma population in contrast, the GFR remained basically unchanged (+0.2 ml/minute [SD 16.4]), and TRVR rose (+566 [SD 16.4]), indicating a defect in their kidneys' ability to handle increased protein loads. Baseline ERPF was lower in patients compared with controls and did not increase with amino acid challenge, in contrast to a 20% increase in controls. When the SSc patients were looked at individually, 15 of the 21 had diminished RFR. They differed from the patients with preserved renal function in respect to shorter time from diagnosis to the study (median 18 months versus 54.5 months), and higher blood pressure (although they remained within the normal range). Statistical analysis showed that the RFR was inversely dependent on basal mean arterial pressure and basal GFR (multiple regression analysis, R2 = 65%, P < 0.0001) in SSc patients, and this relationship did not exist for the control group.
Previous studies have suggested the presence of abnormalities of renal blood flow in patients with SSc. Rivolta et al (2) showed by color flow Doppler ultrasonography significant resistance to flow in the renal arteries of SSc patients with normal creatinine clearance. Studies by Clements et al (3) showed decreased PAH clearance and increased plasma renin activity in patients with SSc, including patients with limited cutaneous scleroderma (lcSSc). Similarly, Kovalchik et al (4) found abnormal serum plasma renin activity that correlated with histopathologic abnormalities on renal biopsy in patients without overt renal disease. Thus, the report by Livi et al (1) confirms earlier studies suggesting renovascular pathology in SSc patients without obvious clinical evidence of renal involvement.
This study demonstrates that SSc patients may have a defective or absent vasodilatory response in the kidney, possibly due to a defect in the endothelial-dependent response to or release of vasoactive substances. The data showed that a larger percentage of patients with diffuse disease compared with limited disease lacked RFR (4 of 5 versus 11 of 16), although the sample sizes are small. More importantly, the RFR of patients with SSc was inversely related to basal levels of both blood pressure and GFR, which may be helpful in predicting which subset of early SSc patients may have underlying renal defects. The finding of impaired renovascular response in many patients with limited scleroderma who are not prone to develop clinical renal disease is noteworthy. Although the authors suggest that a measure of RFR early in disease course may help identify the group at risk of clinical renal disease, their findings in lcSSc would argue against such utility. Interestingly, Clements et al (3) found that abnormal plasma renin activity in asymptomatic SSc patients did not predict the development of renal crisis. Further longitudinal studies are needed to determine the clinical significance of the early defect in renal functional ability demonstrated in this study.
Severe organ involvement in systemic sclerosis with diffuse scleroderma (Arthritis Rheum, 2000)(5)
The clinical presentation of SSc can be diverse both in the extent of skin involvement and the severity of visceral disease. Even in patients with diffuse scleroderma, those with cutaneous disease of the trunk or proximal extremities, there is a range from mild skin involvement to critically ill patients with vital organ damage. Early reports on scleroderma emphasized the progressive nature of the disease and high mortality rates. More recent studies have shown that not all SSc patients inevitably develop internal organ involvement and prognosis is not as dire as it was historically reported. This study was undertaken to determine the rate and timing of severe organ system involvement in patients with diffuse SSc and the natural history of disease in this subset, including their rate of survival.
The authors evaluated all 953 patients with diffuse scleroderma referred to the University of Pittsburgh between January 1, 1972 and December 31, 1995 and followed these patients prospectively for a mean of 10 years. They evaluated these patients for the presence of severe disease involvement and they used specific criteria for each organ system: kidney, renal crisis; cardiac, decompensated heart failure with decreased left ventricular function due to cardiomyopathy, symptomatic pericarditis, or arrhythmia; lung, pulmonary fibrosis visible on chest radiograph with forced vital capacity <55% of predicted (patients with pulmonary hypertension were excluded from this study); gastrointestinal (GI), presence of pseudoobstruction, malabsorption syndrome, or other GI problems requiring hyperalimentation (patients with diarrhea responsive to antibiotics, bloating after meals, or esophageal strictures were not included unless they had 10% weight loss or hospitalization); skin, modified Rodman total skin score >40 from a maximum score of 51.
The mean duration of disease in the group with severe organ involvement was 9.8 years (measured from time of first symptom to last followup contact). Of these, 80% were women and 94% were white. At presentation, 68% of patients had less than 2 years of symptoms and 42% had Raynaud's phenomenon as their first symptom. To confirm the accuracy of their database, they randomly contacted 50 of the 953 patients and verified the presence of severe organ system involvement. They found 97% accuracy in their data; 5 patients developed severe skin involvement after the last skin evaluation and 1 patient developed new onset malabsorption and congestive heart failure.
Of the 953 patients, 656 (69%) with diffuse scleroderma did not develop severe organ involvement as defined above. Of the 397 patients with severe organ damage, skin was the most frequently involved organ system (24%); 19% developed renal crisis, 16% had severe pulmonary fibrosis, 15% had heart involvement, and 8% had GI disease. Of the 233 patients with severe skin involvement, 24% also had kidney disease, 24% had heart involvement, and 15% had lung disease. Of 177 patients with renal crisis, 27% also had heart disease but far less had concomitant lung and GI involvement (11% and 14%, respectively).
The authors then broke the followup period into three 3-year intervals and mapped the timing of the severe organ impairment. In all organ systems, the highest number of cases (steepest portion of the curve) occurred in the first 3 years of followup. For skin, 18% had severe skin involvement by 3 years and this increased by 9% over the next 6 years. The same analysis was done plotting only patients with severe organ disease and its relationship to timing of occurrence. In this category, 80% of patients with severe skin involvement developed it in the first 3 years, and this increased only by 12% in the remaining 6 years. The numbers that developed severe skin disease but later improved is not noted. Although the numbers are smaller, GI involvement occurred in 4% of patients in the first 3 years and this increased to 8% by 9 years. Similar trends were seen for renal disease and slightly less steep but significantly higher rate of involvement in the first 3 years for lung, heart, and GI systems. The authors then further subcategorized the patients with severe organ involvement into early phase of disease (disease <2 years) and later phase of disease. The findings again confirm that renal and skin disease occur more frequently in early disease and mortality from lung and GI disease occurs with higher frequency later in the disease.
The cumulative survival rate (time 0 being time of diagnosis of severe organ involvement) for all patients with scleroderma without severe organ involvement was 72% at 9 years versus 38% in those with any severe organ involvement. Patients with both heart and skin involvement had worse prognoses than those with heart involvement alone, but this was not true for lung, kidney, or the GI tract. With the exception of skin involvement, patients with disease affecting 2 organ systems had similar survival rates to those with each organ system alone. At 9 years, survival in the renal and cardiac groups was 40%, the pulmonary disease group was 30%, and the GI group fared worst at less than 20%. The data on survival of patients with renal crisis changes dramatically when the data is reanalyzed using numbers after ACE inhibitors were used in management. The 9-year survival increases in this group from 40% to 68%. The rate of nonscleroderma death in this population was 38%. The authors also analyzed the rate of death according to duration of disease. Mortality rates were highest in the first 5 years as compared with the next 5 years for patients with severe kidney and heart involvement, but the opposite was true for lung disease, suggesting that the latter continues to progress.
The proportion of patients with renal, cardiac, and lung disease is similar to that seen in other centers. The data suggest that referral bias may have augmented the proportion of patients with severe disease. Among those seen with early disease, within the first 2 years, there was a greater proportion developing severe organ involvement.
The results highlight the likelihood that severe visceral disease tends to occur early in the course of scleroderma. However, an appreciable number of patients develop severe internal organ disease late in the course of scleroderma (particularly cardiac, pulmonary, and GI involvement), where almost half develop severe involvement after 3 years. The assignation of severe lung involvement is based only on forced vital capacity; the proportion that develops pulmonary hypertension as a serious morbidity, either alone or with interstitial lung disease, is not noted.
When evaluating survival in scleroderma, it is important to ascertain that improved survival is not merely reflecting earlier recognition of the disease, i.e., improved survival as an artifact of starting the “clock” earlier. The authors compare survival from first diagnosis of scleroderma for patients without severe organ disease (72% at 9 years) to those with severe organ disease (38% at 9 years). However, once severe organ involvement is documented, subsequent 9-year survival ranges from 15% to 40%. The method of reporting makes comparison to earlier series from the same center difficult (6). Still, the data suggest an improving prognosis in scleroderma compared with series reported 30 years ago. Furthermore, the data suggest that not all the improvement is a result of improved prognosis of renal scleroderma due to availability of ACE inhibitors; subgroups with cardiac and pulmonary disease also appear to have an improved prognosis. Recognition and effective treatment of early visceral disease are the obvious keys to improved survival; identification of effective therapy for pulmonary, cardiac, GI, and skin disease is the challenge.
Long-term outcomes of scleroderma renal crisis (Ann Intern Med, 2000) (7)
Scleroderma renal crisis is a complication of SSc with serious consequences. It is defined as the sudden onset of severe hypertension followed by rapidly progressive renal failure and/or microangiopathic hemolytic anemia. It had been the main cause of mortality in patients with scleroderma until the recent advances in early recognition and treatment of renal involvement with aggressive ACE inhibitor therapy. The authors of this article performed a prospective observational cohort study of a group of patients with scleroderma renal crisis followed at the University of Pittsburgh to determine the long-term renal outcomes of the disease.
The study group for this analysis was selected from all patients with systemic sclerosis referred to the study center between 1979 and 1996. Of 807 patients, 145 had renal crisis and they all received ACE inhibitor therapy. This group was then followed regularly with 95% accountability over the next 5–10 years. The study group was divided into 4 categories for analysis: 1) those who did not require dialysis within the first year after renal crisis, 2) those who required temporary dialysis, 3) those who required permanent dialysis, and 4) those who died within the first 6 months of the renal crisis. They evaluated the methods and complications of dialysis and compared them with the numbers in the literature for the general population receiving dialysis.
Scleroderma renal crisis was seen early in the course of SSc, with three-fourths of cases occurring within the first 4 years of disease onset (median 2.4 years). Of the study population, 75% were women, 92% were white, and the mean age of onset of renal crisis was 50 years. Diffuse skin disease was seen in 88% of patients; 12% had less extensive skin involvement but had other findings suggestive of diffuse disease (tendon friction rubs, antitopoisomerase I antibody, and anti-RNA polymerase III antibody ); only 4 patients had lcSSc. New-onset diastolic hypertension (mean peak blood pressure, 178/102 mm Hg) as a manifestation of renal crisis was seen in 85% of patients. At the time of initial evaluation, 90% of patients had an elevated creatinine, with a mean of 2.8 mg/dl and 45% had microangiopathic hemolytic anemia.
The breakdown of patients into the 4 outcome categories was as follows: there were 55 patients (38%) who did not require dialysis, 34 patients (23%) received temporary dialysis (2–18 months, mean 8 months), 28 patients (19%) received permanent dialysis, and 28 patients died early (mean of 3 months after onset of renal crisis). In the group who never required dialysis, the serum creatinine peaked at 3.8 mg/dl and fell to 1.8 mg/dl at the end of 7.1 years. Only 2 patients from this group had slow deterioration of renal function requiring dialysis 4–6 years later. All patients continued ACE inhibitor therapy. One patient developed recurrence of malignant hypertension and developed renal failure despite ACE inhibitor therapy.
In the temporary dialysis group, mean creatinine concentration was 2.7 mg/dl at the time of discontinuation of dialysis and continued to improve to 2.2 mg/dl at the end of 6.1 years. Only 1 patient who was noncompliant with the ACE inhibitor therapy returned to dialysis. Of patients with permanent dialysis, 72% received hemodialysis while the rest used peritoneal dialysis. Six patients received kidney transplants after 4 or more years of dialysis and none had recurrence of renal crisis. Of patients in the permanent dialysis group, 75% continued taking ACE inhibitors (including the 4 who received transplants).
The early mortality group had more men (33% versus 16% men in the other groups), was older at diagnosis (54 versus 46 years), and had higher initial creatinine levels. More patients in this group had scleroderma myocardial disease (43% versus 5%) than those in no dialysis or temporary dialysis groups. The cause of death included problems with dialysis, infection, and general multiorgan system failure.
Overall survival rates in the groups that did not require dialysis and that required temporary dialysis were equivalent to SSc patients without renal crisis, with a 5-year cumulative survival rate of 90% and an 8-year survival of 80–85%. Those on permanent dialysis predictably had significantly worse survival rates, with 5-year survival at 40% and 8-year survival at about 25%. The dialysis-related complication rates in scleroderma patients (0.44 per patient-year in the permanent dialysis group and 0.58 per patient-year in the temporary dialysis group) were similar to those reported in the literature of dialysis patients without scleroderma. The rate of graft thrombosis was similar in the 2 groups (with and without scleroderma), and only 1 patient with severe Raynaud's phenomenon required removal of access graft due to a severe new digital-tip ulcer. The patients using peritoneal dialysis also had similar rates of complication, such as catheter infection, obstruction, and peritonitis, as those receiving peritoneal dialysis without scleroderma.
As reported by Steen et al (9), ACE inhibitor therapy has revolutionized the treatment of renal crisis in patients with scleroderma. With the continued use of ACE inhibitors in this population, this article shows that most patients who never require dialysis or those who require temporary dialysis can have excellent renal outcomes with no difference in mortality rate versus scleroderma patients without renal crisis. It also clearly demonstrates that the majority of these patients will not go on to develop worsening renal function nor will most ever develop renal crisis again. Furthermore, 37% of patients who initially required dialysis were able to come off dialysis and their long-term outcomes were good.
The overall prevalence of renal crisis in SSc and its effect on survival cannot be ascertained from this study. Because the authors' program is a referral center for SSc, the population in this study may be biased towards more severe renal disease and potentially worse outcomes. In previous studies, patients whose ACE inhibitors were started when the creatinine was less than 3 mg/dl were unlikely to require dialysis (9). It is unclear how many patients in this cohort had untreated or inadequately treated renal crisis before first being seen. Conversely, patients treated early and/or with good responses to ACE inhibitors might be less likely to be referred. These factors would tend to make overall results in the study cohort worse than in the general population of SSc patients. It is thus noteworthy that even in this select population, outcomes were good.
The current standard of practice in SSc is that ACE inhibitor therapy should be started as soon as scleroderma renal crisis is diagnosed. Nonetheless, a significant number of patients with SSc develop life-threatening renal crisis and many require dialysis at, or shortly after, presentation. It is reassuring to learn from this study that with appropriate diagnosis and aggressive treatment, these patients can have good outcomes in terms of survival and becoming dialysis-free.
Bosentan therapy for pulmonary arterial hypertension (N Engl J Med, 2002) (10)
There are 2 predominant forms of lung disease seen in SSc: interstitial lung disease (inflammatory alveolitis leading to interstitial fibrosis) and pulmonary hypertension (PHT). Pulmonary hypertension in scleroderma is insidious in onset and patients are asymptomatic until much of the lung function is lost. The true incidence of PHT in SSc is unknown, but it is estimated to affect 25–50% of patients with lcSSc (CREST [calcinosis, Raynaud's phenomenon, esophageal dysmotility, sclerodactyly, telangiectasias] syndrome) and is thought to be the leading cause of death in this subset. The onset of PHT is often years after onset of the disease and is “primary” vascular disease. In patients with diffuse scleroderma it more commonly occurs along with interstitial lung disease.
Until recently, treatment options for PHT were limited. Calcium channel blockers are almost uniformly ineffective in PHT in the setting of scleroderma. Several years ago, continuous infusion of epoprostenol was shown to be an effective treatment modality in primary idiopathic PHT and in scleroderma-associated PHT (11). However, this is a very expensive and cumbersome therapy. Because of its short half-life, epoprostenol must be given as a continuous infusion via a central line, which is not only cumbersome and hazardous but also very expensive. As a result, its use has been reserved for patients with advanced disease. A subcutaneous prostacyclin, treprostinil, has been approved recently; it also is very expensive and must be given by continuous infusion (12).
Recent investigations have revealed that endothelin-1, a potent endogenous vasoconstrictor, is one of the important contributors to pulmonary hypertension (13). It has been found to be overexpressed in plasma and lung tissues of patients with both idiopathic pulmonary hypertension and in patients with SSc. Bosentan, an endothelin receptor antagonist (to both endothelin receptors ETA and ETB), was found in preliminary studies to be safe at an oral dosage of 125 mg twice a day and patients demonstrated improved exercise capacity. The summarized study (10) is a multicenter trial of bosentan at dosages of 125 mg twice a day and 250 mg twice a day evaluating its efficacy on exercise capacity in patients with both idiopathic pulmonary hypertension and PHT secondary to connective tissue diseases, including systemic lupus erythematosis and SSc.
This was a double-blind, randomized, placebo-controlled trial with 213 patients recruited from 27 centers around the world, including Europe, North America, Israel, and Australia. To be eligible for inclusion in the study, patients had to be World Health Organization (WHO) class III (dyspnea with mild exertion) and IV (dyspnea at rest) with the caveat that patients in class IV must be sufficiently clinically stable to ethically allow participation in a placebo-controlled trial. Patients were excluded if they had been taking any therapy for PAH within the past month before screening, if they were designated to receive epoprostenol within 3 months before screening, and/or if they were taking glyburide or cyclosporin.
The patients were randomized into the placebo arm (69 patients) versus 62.5 mg of bosentan twice daily for 4 weeks, then the bosentan group was split into 125 mg (74 patients) versus 250 mg (70 patients) twice daily for 12 additional weeks. This portion was designated period 1. Period 2 included only patients randomized within the first 2 months of the study, and these patients (48) completed an additional 12 weeks of treatment. Then all patients were eligible for open-label study of the drug.
The outcome measure designated for period 1 was the change in 6-minute walk time from baseline to the end of week 16. The secondary measures included the change in the Borg dyspnea index (measure of a person's perceived breathlessness on a scale from 0 to 10 with 10 being the worst), a change in the WHO functional class, and the duration of time to clinical worsening from the time of randomization (defined as combined end point of death, lung transplant, hospitalization for PAH, lack of clinical improvement or worsening leading to discontinuation, need for epoprostenol therapy, or atrial septostomy). The data from period 2 of the study was used to evaluate the drug's safety and efficacy. All data were analyzed on an intent to treat basis.
The baseline characteristics were similar between the groups. Primary PHT was more common than that related to connective tissue disease. The duration of disease since diagnosis was 28 months in the placebo group versus 30 in the bosentan group. At baseline, the 6-minute walk distance in the placebo group was 344 meters versus 330 meters in the bosentan group, the dyspnea scale was similar, but there were only 4 patients in the placebo group in WHO class IV versus 6 and 8 in the 2 bosentan groups.
At the end of 16 weeks, the 6-minute walk time for the bosentan group increased by 36 meters, while it decreased by 8 meters in the placebo group, with the mean difference of 44 meters (95% confidence interval [95% CI] 21, 67; P < 0.001). Although the distance walked was greater in the 250 mg group compared with the 125 mg group, this difference was not statistically significant. The data were then further analyzed by subgroups. In the patients with primary PHT, the 6-minute walk distance improved by 46 meters (whereas it deteriorated by 5 meters in the placebo group); in patients with PHT secondary to scleroderma, bosentan prevented deterioration of their 6-minute walk distance compared with a decrease by 40 meters in their counterparts in the placebo group.
The bosentan group also improved in secondary outcome measures, including the Borg dyspnea index (mean effect of −0.6 for bosentan with 95% CI −1.2, −0.1). Many patients in the bosentan group also improved their WHO classification by the end of 16 weeks; 38% of patients in the 125 mg group and 34% of patients in the 250 mg group improved from class III to II (dyspnea with moderate exertion), and 3% in the 125 mg group and 1% in the 250 mg group improved to class I, whereas 28% of patients in the placebo group improved to class II and none to class I. Bosentan also increased time to clinical worsening (defined above) significantly compared with placebo, although there was no difference between the 2 dosages. The difference in the 2 groups was detected at week 16 and persisted through period 2, which ended at 28 weeks.
The number of adverse events was similar between the placebo and bosentan group, with the exception of abnormal liver function test (LFT). Nine patients (6%) from the 2 bosentan groups discontinued the study due to adverse events, whereas 5 patients (7%) dropped out of the placebo group. Three patients in the bosentan group withdrew from the study due to LFT abnormalities (most frequent adverse event in the bosentan group). This was found to be dose-dependent: 2 patients (3%) in the 125 mg group had transaminases 8 times the upper limit of normal and 5 patients (7%) in the 250 mg group had similar elevations. Six patients in the placebo group versus 2 patients in the bosentan group withdrew due to clinical worsening of symptoms of PAH and syncope. There were 3 deaths and they all occurred during period 1. Two patients in the placebo group died of worsening PAH and 1 patient in the 125 mg bosentan group died of heart failure. Three patients in the 250 mg bosentan group died within 4 weeks of withdrawing (pneumonia) or completion of the study (pulmonary hemorrhage during the open-label extension). Overall, 14% of bosentan-treated patients developed LFT abnormalities greater than 3 times the upper limit of normal.
The therapeutic options for patients with PHT are limited. For many patients, their lives are severely limited by dyspnea. The primary outcome variable in this study is an important functional result. Although an improvement of 36 meters in the 6-minute walk time may seem small, it may have important functional consequences for activities of daily living. Although the walk distance did not improve for patients with PHT secondary to SSc, it did prevent further deterioration compared with placebo, which is significant. Patients in the bosentan group also had a major reduction in the time to clinical worsening within the study period, suggesting that it may slow progression of disease, but longer studies will be needed to confirm this trend.
Although this study was not designed to show improved survival, the authors point out that the 6-minute walk distance has been shown to be an independent predictor of mortality and the patients in the epoprostenol study had both improved walk distance and improved survival. Because there were significantly more substantially elevated transaminases in the bosentan 125 mg group, and because there was not a statistically significant difference in walk time between the 2 doses, the recommendation by the authors is to use the 250 mg twice daily dosing with careful monitoring of LFT.
Although bosentan will not be the ultimate answer to PAH, it is an important option for patients, especially in the US where the only other option is continuous infusion of epoprostenal. Bosentan's efficacy in well-established disease is not as dramatic as the improvement seen with epoprostenol, but for patients with mild to moderate PAH, it may provide improved quality of life. Additional studies and postmarketing data may help to determine whether bosentan will have an impact on slowing the progression of disease and affecting mortality rates.
Histopathologic subsets of fibrosing alveolitis in patients with systemic sclerosis and their relationship to outcome (Am J Respir Crit Care Med, 2002) (14)
Interstitial lung disease with subsequent pulmonary fibrosis is a significant cause of morbidity and mortality in patients with SSc. Although the exact mechanism of disease is unclear, there is an influx of inflammatory cells into the interstitium and alveolar spaces, leading to destruction of the normal architecture. To ascertain the degree of inflammation, a bronchoalveolar lavage (BAL) is frequently performed in these patients. The finding on the BAL may affect decisions regarding therapy; if increased numbers of inflammatory cells are found, patients may require aggressive therapy with immunosuppressive agents, such as cyclophosphamide, which has been shown to improve pulmonary function as well as survival in this group (15).
The authors of this present study have published previous articles demonstrating better prognosis in patients with interstitial lung disease secondary to SSc than those with idiopathic pulmonary fibrosis (16, 17). More recently, they compared BAL cellularity between the 2 groups and found that patients with cryptogenic or idiopathic fibrosing alveolitis had higher levels of both neutrophils and eosinophils in the BAL fluid, but the eosinophil levels were ultimately found to be linked to more extensive lung disease (18). In the past 5 years, interstitial lung disease has been further categorized into specific histologic patterns: nonspecific interstitial pneumonia (NSIP) and usual interstitial pneumonia (UIP). This study was undertaken to incorporate the histopathologic findings on the biopsies of SSc patients with interstitial lung disease with other parameters of disease extent, including pulmonary function test (PFT) results, BAL cellularity, and clinical features, in predicting prognosis of these patients.
This was a retrospective study done on subjects referred to Royal Brompton Hospital in London from January 1985 to December 1997 with followup until January 31, 2001 or death of the patient. Of the 476 patients with SSc, lung biopsies were performed on 94 patients. Lung biopsies were done on all patients with what was deemed by physicians as clinically significant interstitial lung disease. Of the 94 patients with biopsies performed, clinical information and histopathologic data were available on 80 patients. In all patients who underwent a computed tomography (CT) scan prior to biopsy, the scan revealed pulmonary fibrosis and in the 2 patients without a CT scan, the chest radiograph was consistent with pulmonary fibrosis. This study included patients with both limited and diffuse scleroderma but excluded patients with overlap syndromes. The patients were treated with prednisone with or without immunosuppressive agents, including azathioprine and cyclophosphamide or D-penicillamine.
The histologic specimens were evaluated independently by 2 histopathologists who were blinded to patient information. The specimen was determined to be UIP if it revealed interstitial lung disease at varying stages, ranging from fibroblastic proliferation with minimal collagen deposition to severe fibrosis. In NSIP, the damage was homogeneous, therefore, they were further categorized to cellular NSIP or fibrotic NSIP. A fourth category of end-stage lung (ESL) was designated for samples with complete loss of architecture; and finally, there was a miscellaneous category (that was not included in the analysis), which included patients with respiratory bronchiolitis interstitial lung disease (4 patients), sarcoidosis (1 patient), and organizing pneumonia (1 patient). Excluding this last category left 74 patients for analysis.
The results revealed that the agreement between the 2 pathologists were good but not excellent (kappa coefficient of agreement was 0.72). Differences were discussed by the pathologists and resolved. The most common histology was NSIP (62 of 80; 77.5%) and within the NSIP category, 24% was cellular NSIP and 76% was fibrotic NSIP. UIP was seen in 6 patients and ESL was found in 6 patients. Because histopathologically ESL is more similar to UIP, these 2 groups were lumped together in the clinical analysis.
The median age of the group was 46 years (range 23–69 years) and the majority were women (61/74). A larger percentage of patients in this study had limited scleroderma than diffuse disease (70% and 30%, respectively), which is unusual because literature reports that interstitial lung disease occurs predominantly in patients with diffuse scleroderma. Pulmonary hypertension was diagnosed in 21% of patients by echocardiogram. The mean (± SD) duration of respiratory symptoms before referral to their center was 13 months (±13 months). Clinical symptoms of dyspnea on exertion (DOE), chronic cough, or wheeze were found in 89% of patients and on clinical exam, 85% had bilateral basal rales. These findings are consistent with those reported by Witt et al (19) showing DOE in 78% of patients, cough in 55%, and late inspiratory crackles at the bases in 47% of patients in his cohort.
Upon initial evaluation, 51 of the 74 patients had abnormal PFTs; 42 with restrictive disease, 2 with obstructive disease, and 7 with a mixed pattern. Diffusion capacity (DLCO) was normal in only 2 patients. Overall, patients with UIP/ESL had lower DLCOlevels than patients with NSIP, but forced vital capacity (FVC) levels did not differ significantly between the 2 groups. The initial DLCOand FVC levels were also not significantly different between the 2 NSIP groups. Of the 67 patients whose BAL information is available, the eosinophil level was higher in NSIP than in UIP/ESL (P = 0.02). The neutrophil and lymphocyte levels were not significantly different and the neutrophil and eosinophil counts were not significantly different between the 2 NSIP groups. Lymphocyte count was higher in cellular NSIP than fibrotic NSIP.
The followup for these patients ranged from 1 to 16 years and during that time, 22 (30% of total) patients died: 16 patients with NSIP (26% of those with NSIP) and 6 with UIP/ESL (50% of those with UPI/ESL) (4 of 6 with UIP and 2 of 6 with ESL). Lung cancer developed in 4 patients: adenocarcinoma in 3 and small cell CA in 1. Mortality rate was increased in patients with lower FVC and DLCO at initial presentation. It did not vary with sex, age, smoking status, BAL findings, and whether the scleroderma was limited versus diffuse. Surprisingly, there was no significant difference in survival rates between NSIP and UIP/ESL, although a trend in favor of better survival in NSIP was found. Published data report increased survival rates in NSIP versus UIP in idiopathic cases (20). The 5-year survival rate in this study was 90% in NSIP and 82% in UIP/ESL and although the 10-year survival shows a significant divergence between the survival curves (69% in NSIP and 29% in UIP/ESL), this was not found to be statistically significant because of the small number of patients in the latter group. Upon further subgroup analysis of NSIP alone, there was no difference in survival between cellular NSIP and fibrotic NSIP. Within the NSIP group, there was a statistically significant increase in mortality in patients with lower DLCO level at initial assessment and in patients with higher levels of eosinophils in the BAL fluid. The authors also compared serial PFTs in patients with NSIP. The data show minimal change in FVC or DLCO at 1 and 3 years in patients undergoing treatment. Although changes in FVC at 1 and 3 years and changes in DLCO at 1 year were not linked to improved survival, a change in DLCO at 3 years did affect survival; improvement of the DLCO was associated with increased survival and decrease in DLCO resulted in decreased survival rates. It is unclear if the increased mortality in the group with worsening DLCO at 3 years was secondary to advancing interstitial disease or pulmonary hypertension because repeat echocardiograms were not done at 3 years.
The most significant finding in this study was the high prevalence of NSIP within this population of SSc patients compared with a higher proportion of UIP in patients with idiopathic pulmonary fibrosis reported in the literature. There are some inherent problems with histopathologic studies: selection bias of patients and sample selection. The authors address the selection bias issue by declaring that biopsies were offered to all patients who were determined to have “significant interstitial lung disease” and not just to patients with specific patterns found on CT scans, but what criteria were used to determine significance is not reported. They also specify that patients with severe disease who were unable to tolerate a biopsy and those who refused biopsy were excluded but the numbers in this group was small, and they feel that this would not have changed the final results. It is not described in the text how the area of biopsy was chosen. Some pathologists have reported the presence of both NSIP and UIP in a patient with interstitial lung disease and therefore, if specific criteria were used to select the area for biopsy, it may have skewed it in favor of one pathology versus another.
The other surprising finding from this study was that there was no significant difference in survival between NSIP and UIP/ESL patients. Multiple articles have shown improved survival in patients with idiopathic NSIP when compared with idiopathic UIP. The authors state that the results in this study on survival must be interpreted with caution because of the small numbers of patients in the UIP/ESL group and because of the difficulty in assessing how much the lung disease contributed to mortality. Although the 4 patients that developed lung cancer died of their malignancy, the cause of death was not identified in the other 18 patients. The lumping of the ESL group with UIP may also be skewing results. Histologically, ESL is more similar to UIP than NSIP, but because the natural history of ESL is unknown, improved survival in ESL may have be contributing to improved survival in the UIP/ESL group. There were many more patients with fibrotic NSIP than cellular NSIP in this study. Because of the small number of patients in the cellular NSIP group (15 versus 47), the study may not have been able to detect a difference in survival between the 2 groups (data not shown). If fibrotic NSIP has survival rates closer to that of UIP, this may explain the absence of difference in survival between UIP/ESL and NSIP groups.
Interstitial lung disease is a devastating complication of SSc with high mortality and morbidity. It is interesting to find a higher prevalence of NSIP in scleroderma patients compared with higher percentage of UIP in idiopathic fibrosing alveolitis, but further studies will be needed to evaluate prognostic differences in patients with the varying histopathologic findings. It is encouraging to see the high survival rates and the lack of functional decline in treated patients with NSIP in this study.
Anti-TGF-β treatment prevents skin and lung fibrosis in murine sclerodermatous graft-versus-host disease: a model for human scleroderma (J Immunol, 1999) (21)
Although much has been learned about the pathogenesis of SSc in recent years, the ultimate cause of disease is still unclear. Through the work on inbred Choctaw Indians, in whom the incidence of scleroderma is many-fold higher than the general population, it is clear that the disease has a genetic component. But other publications have also suggested contribution by environmental agents (including infections) and microchimerism (the persistence in the mother of fetal stem cells that migrated across the placenta during pregnancy or, conversely, maternal cells that persist in the child long after birth). Microchimerism is found with high prevalence in autoimmune thyroid disease as well as in SSc, and interestingly, pathogenesis of these diseases appears to be similar to that of graft-versus-host disease (GVHD) found in patients after bone marrow transplantation (BMT).
This study evaluates a murine sclerodermatous GVHD model to assess its value as a prototype of human SSc. It has been established that in human SSc, there is an increased production of collagen at both mRNA level and at the protein level (22). This increase appears to be linked to increased levels of transforming growth factor β1 (TGF-β1), which is a known powerful stimulator of collagen synthesis, and in the pathogenesis of both pulmonary and skin fibrosis in SSc. The authors of this study set out to demonstrate that this murine model of GVHD is not only phenotypically similar to human SSc with thickening of the skin and pulmonary fibrosis, but that there is upregulation of collagen and TGF-β, similar to the human SSc. They also examined whether blocking of TGF-β receptors with antibody can prevent systemic fibrosis.
The mouse model of GVHD was created by irradiating 7–8-week-old BALB/c strain of mice and 6 hours later, transplanting donor spleen and bone marrow cells; BALB/c spleen and bone marrow cells in the control group (syngenic BMT) and B10.D2 (which differs from the recipient strain at the minor histocompatibility loci) spleen and bone marrow cells in the GVHD model. The animals were killed at days 7, 14, 21, 38, 49, and 75 post-BMT and skin was obtained for RNA extraction and for histologic study. Real time polymerase chain reaction (RT PCR) was done to measure the upregulation of TGF-β1 mRNA and proα1(I) collagen mRNA was quantitated using RNase protection assays. Specific antibodies were used for immunostaining to identify CD11b/CD18 monocytes/macrophages to confirm their presence at sites of TGF-β upregulation, which the authors propose as being activated by host-reactive T cells with subsequent production of increased levels of TGF-β.
The mice receiving syngenic BMT did not develop GVHD. The mice receiving mismatched bone marrow did develop GVHD, demonstrated histologically as apoptotic keratinocytes and satellitosis in the epithelium of the tongue and skin. Mild skin thickening was noted by day 14 after BMT but significant skin thickening (increase in total dermal area of 40% over controls) did not occur until day 21 and remained steady up to day 76. The affected mice also demonstrated histologic differences in lung tissue, with decreased alveolar spaces by about 30% at day 21 post-BMT. RT PCR demonstrated about a 3-fold increase in TGF-β1 mRNA in the GVHD mice versus controls, and it occurred as early as day 6 and decreased slightly to about 2-fold by day 21. The collagen mRNA rose steadily with 1.8-fold increase by day 21 post-BMT compared with control animals, peaking to a 15-fold increase by day 38, then falling back down to 1.5-fold increase by day 75 (the fold difference reported after correcting against a housekeeping gene, β-actin). Immunostaining demonstrated significant influx of CD11b+ mononuclear cells by day 21 in animals with GVHD, and they were highly concentrated in the deep dermis where fibrosis occurs.
To test the hypothesis that increased fibrosis of scleroderma and GVHD occurs as a consequence of increased collagen deposition secondary to upregulation of TGF-β, an antibody to TGF-β was infused into the mice on day 1 and 6 post-BMT and evaluated on day 21. There was no difference in the rate of success of the BMT, but the animals that received the antibody developed neither skin thickening nor pulmonary fibrosis. There was no change in the skin or lung of control animals that received the antibody.
There are other murine models of scleroderma, including the tight skin (Task) mouse, which has a mutant fibrillin gene. This is the same gene that is closely linked to the scleroderma disease found in the Choctaw Indian population. Although these mice have thick skin, they do not develop vascular changes or interstitial lung disease of humans. This current model of GVHD is promising as an animal model of SSc, with similar biochemical and phenotypic features to that seen in human disease, and may be helpful in the future for further advancing our knowledge of this disease.
The pathogenesis of SSc is obviously complex but with recent advances in understanding of the disease, new targets of therapy are being uncovered. The evidence from this study, as well as a wealth of evidence on involvement of TGF-β in fibrosis, suggest that TGF-β is a rational target for therapy in scleroderma. Phase 1 safety trials are currently underway to evaluate the safety of TGF-β antibody in patients with SSc.