A comparison between anti-Th/To– and anticentromere antibody–positive systemic sclerosis patients with limited cutaneous involvement

Authors


Abstract

Objective

To compare anti-Th/To–positive and anticentromere antibody (ACA)–positive patients with limited cutaneous systemic sclerosis (lcSSc).

Methods

We reviewed the medical records of 107 anti-Th/To–positive patients and 365 ACA-positive patients who were first evaluated during 1985–2000. ACA was detected by indirect immunofluorescence on HEp-2 cell substrate, and anti-Th/To was detected by RNA immunoprecipitation with K562 cell extracts. Patients were included if they had a clinical diagnosis of lcSSc, and excluded if they had >1 SSc-associated serum autoantibody.

Results

The final study groups comprised 87 lcSSc patients with anti-Th/To antibodies and 306 with ACAs. Anti-Th/To–positive patients were younger (P < 0.04) and had a shorter disease duration at their first evaluation (P < 0.003). Patients with anti-Th/To antibodies had more subtle skin changes, less severe vascular involvement, and less frequent distal esophageal hypomotility. Both groups had a higher frequency of “intrinsic” pulmonary hypertension than has been previously reported in the literature (28% and 19% of anti-Th/To–positive and ACA-positive patients, respectively), perhaps due to referral bias. Patients in the anti-Th/To group more often had radiographic evidence of interstitial lung disease (48% versus 13% of the ACA group; P < 0.0001). Scleroderma renal crisis was uncommon (4 cases), but occurred exclusively in the anti-Th/To group. Survival among the anti-Th/To–positive patients was reduced compared with that in the ACA group (P < 0.02).

Conclusion

Patients with anti-Th/To and those with ACA most often develop lcSSc and have a high frequency of “intrinsic” pulmonary hypertension. Compared with the ACA patients, anti-Th/To lcSSc patients have more subtle cutaneous, vascular, and gastrointestinal involvement, but more often have certain features typically seen in diffuse scleroderma, such as pulmonary fibrosis and scleroderma renal crisis, as well as reduced survival.

Systemic sclerosis (SSc) is characterized by Raynaud's phenomenon, thickening of the skin, and involvement of various internal organs (1). The disease is highly variable in its presentation, ranging from limited cutaneous involvement (lcSSc) in which skin thickening is restricted to the fingers, hands, and/or face, with less serious internal organ involvement, to diffuse cutaneous involvement (dcSSc) in which skin changes are widespread and often rapidly progressive, with earlier and more serious visceral complications (1, 2).

Antinuclear antibodies (ANAs) have been found in >90% of sera from patients with SSc (3). Several types of ANAs are recognized to have diagnostic and prognostic significance (1). Persons with lcSSc frequently have anticentromere antibodies (ACA) or anti-Th/To (4) antibodies. Individuals with dcSSc most frequently have anti–topoisomerase I or anti–RNA polymerase I and III antibodies (5, 6). Patients with SSc in overlap tend to have anti–U1 RNP (7), anti-PM/Scl (8, 9), or anti–U3 RNP antibodies (10). The proportion of SSc patients having any 1 of these 7 SSc-associated autoantibodies is ∼85% (1).

In 1990, investigators in our group compared 14 anti-Th/To–positive lcSSc patients with a large group of anti-Th/To–negative lcSSc patients and showed that the patients with anti-Th/To antibody more often had puffy fingers, small bowel involvement, hypothyroidism, and poor survival (4). Two limitations of this and other prior reports on anti-Th/To antibody are the small number of patients evaluated and the relatively brief period of patient followup. The purpose of the present study was to present the longitudinal clinical and survival data on a large series of anti-Th/To–positive lcSSc patients, and to compare them with a parallel series of ACA-positive lcSSc patients.

PATIENTS AND METHODS

Patients.

All anti-Th/To antibody–positive and ACA-positive patients first seen at the University of Pittsburgh during 1985–2000 were eligible for inclusion in the study. Initial and all followup medical records available were reviewed to determine the most recent diagnosis and disease manifestations.

Laboratory methods.

All sera were diluted 1:40 in 10 mM phosphate buffered saline, pH 7.3, and tested for ACA by indirect immunofluorescence (IIF) performed on HEp-2 substrate (Diasorin, Stillwater, MN). Specimens exhibiting a nucleolar pattern on IIF were evaluated for anti-Th/To by RNA immunoprecipitation as previously described (4). Briefly, a 20-μl serum sample was bound overnight at 4°C to 2 mg protein A–Sepharose CL-4B beads (Amersham Biosciences, Piscataway, NJ), washed 3 times with immunoprecipitation buffer (10 mM Tris HCl, pH 8.0, 500 mM NaCl, 0.1% Igepal), and incubated for 2 hours at 4°C with unlabeled extract from ∼6 × 106 rapidly dividing K562 cells. The beads were washed 3 times with NET-2 buffer (50 mM Tris HCl, pH 7.4, 150 mM NaCl, 0.05% Igepal) and suspended in 300 μl NET-2, 30 μl 10% sodium dodecyl sulfate, 30 μl 3M NaAc, and 2 μl of glycogen per sample. The RNA was phenol extracted and alcohol precipitated. The resultant RNA pellet was suspended in RNA sample buffer and electrophoresed at 400V on a standard size 8% urea–polyacrylamide electrophoresis gel. The gel was then silver stained for RNA visualization.

Study design.

The final study groups included only patients with the diagnosis of lcSSc and having either anti-Th/To or ACA. Individuals with diagnoses other than lcSSc or who had >1 SSc-associated serum autoantibody were excluded. We compared the demographic characteristics, organ system involvements, other clinical features, and survival in these 2 patient groups.

Definitions of organ system involvement.

SSc organ involvement was considered to be present if predefined clinical features in the following 8 organ systems were observed during the course of the illness and were not attributable to other diseases: 1) vascular (Raynaud's phenomenon or any 1 of digital pitting scars, digital tip ulceration, or digital gangrene); 2) cutaneous (skin thickening either absent [SSc without scleroderma] or restricted to regions distal to the elbows and knees, i.e., not affecting the upper arms, thighs, anterior chest, or abdomen; facial skin thickening was acceptable [assessed by the modified Rodnan scoring method (11)]); 3) articular (any 1 of joint swelling or joint contractures, carpal tunnel syndrome, palpable tendon friction rubs, or either joint space narrowing or erosions on radiograph); 4) muscular (proximal muscle weakness on physical examination plus either elevated serum creatine kinase levels or myopathic changes on electromyogram); 5) gastrointestinal (any 1 of distal esophageal dysmotility [by esophagram or motility study], evidence of hypomotility of the duodenum or small intestine, colonic sacculations, or malabsorption syndrome); 6) pulmonary (any 1 of restrictive lung disease [forced vital capacity (FVC) <70% of predicted plus forced expiratory volume in 1 second/FVC >80%], diffusing capacity for carbon monoxide <65% of predicted, pulmonary fibrosis on chest radiograph, “intrinsic” pulmonary hypertension defined as either a mean pulmonary artery pressure of >30 mm Hg on cardiac catheterization or echocardiographic evidence of increased pulmonary artery pressure, pleuritic chest pain plus a pleural friction rub or pleural effusion); 7) cardiac (any 1 of estimated left ventricular ejection fraction <45% or left-sided congestive heart failure, pericarditis [pericardial pain and either a pericardial friction rub or pericardial effusion], arrhythmia requiring treatment, or complete heart block); 8) renal (clinical evidence of scleroderma renal crisis, defined as the abrupt onset of accelerated arterial hypertension or rapidly progressive oliguric renal failure).

Statistical analysis.

Student's t-test was used to detect significant differences between distributions of continuous data. Chi-square analysis was used to determine significant differences between sets of categorical data, with Fisher's exact test being used when appropriate. Because multiple comparisons were made, we defined statistically significant as a P value less than 0.01. The Mantel-Haenszel test was used to detect significant differences between survival curves.

RESULTS

Study group.

Patients with >1 SSc-associated serum autoantibody were excluded. One patient had both the anti-Th/To antibody and ACA. Eight anti-Th/To–positive patients had 2 SSc-associated serum autoantibodies, including 4 patients with anti–U3 RNP, 2 with anti–U1 RNP, 1 with anti-PM/Scl, and 1 with anti–topoisomerase I. Ten ACA-positive patients were similarly excluded because 6 had anti–topoisomerase I, 3 had anti–U1 RNP, and 1 had anti–U3 RNP antibody.

The remaining patients comprised 107 with anti-Th/To antibodies and 365 with ACAs (Table 1). Ninety-six (90%) of the anti-Th/To antibody–positive patients and 336 (92%) of the ACA-postive patients had SSc. Isolated Raynaud's phenomenon was the clinical diagnosis in 7% of both groups. The 3 patients with dermatomyositis/polymyositis, 3 with Sjögren's syndrome, 1 with interstitial lung disease, and 1 with autoimmune hepatitis have not developed any other autoimmune disease after >10 years of followup. In 2000, serum was again obtained from the latter 2 patients who were initially positive for anti-Th/To, and anti-Th/To antibodies were again demonstrated. After further excluding patients with dcSSc and unclassified patients, the final study groups for comparison contained 87 anti-Th/To–positive and 306 ACA-positive lcSSc patients (Table 1).

Table 1. Classification of anti-Th/To– and ACA-positive patients by clinical diagnosis*
Patient groupAnti-Th/To (n = 107)ACA (n = 365)
  • *

    Values are the number (%) of patients. ACA = anticentromere antibody.

  • The final group comprised those with limited cutaneous systemic sclerosis (SSc) after excluding those with >1 SSc-associated autoantibody.

Systemic sclerosis96 (90)336 (92)
 Diffuse cutaneous522
 Limited cutaneous89313
 Unclassified21
Isolated Raynaud's phenomenon8 (7)24 (7)
Dermatomyositis/polymyositis1 (1)2 (1)
Sjögren's syndrome03 (1)
Autoimmune hepatitis1 (1)0
Interstitial lung disease1 (1)0
Final study group87 (81)306 (84)

Demographic characteristics.

The demographic characteristics of the final study group patients are shown in Table 2. The majority of the patients were white and female. There were more white patients and female patients in the ACA group. Anti-Th/To–positive patients presented to our medical center at a younger mean age (51.6 years versus 55.2 years) and with a significantly shorter mean disease duration (10.1 years versus 14.2 years) than that of ACA patients.

Table 2. Demographic characteristics of the final study group patients*
 Anti-Th/To (n = 87)ACA (n = 306)P
  • *

    See Table 1 for definitions.

  • Higher alpha level chosen for statistical significance (P < 0.01) because of multiple comparisons.

% white93980.0248
% female80910.0525
% with education 12+ years90910.810
Mean age at symptom onset, years41.541.00.789
Mean age at diagnosis of SSc, years48.750.80.193
Mean age at first Pittsburgh evaluation, years51.655.20.020
Duration from symptom onset to first physician diagnosis of SSc, years7.29.80.028
Duration from symptom onset to first Pittsburgh evaluation, years10.114.20.002

Organ system involvement.

Table 3 shows the details of organ system involvement and other clinical features. Vascular disease occurred in 99% of patients in each group, and Raynaud's phenomenon was almost universal. However, patients in the ACA group had more severe vascular involvement, including significantly higher frequencies of digital pitting scars, digital tip ulcers, and digital gangrene. Amputations, specifically upper extremity amputations, were reported only in the ACA group.

Table 3. Organ system involvement and other clinical features*
 Anti-Th/To (n = 87)ACA (n = 306)P
  • *

    Except where indicated otherwise, values are the no. (%) of patients. DLCO = diffusing capacity for carbon monoxide; PHT = pulmonary hypertension (see Table 1 for other definitions).

  • Higher alpha level chosen for statistical significance (P < 0.01) because of multiple comparisons.

  • Mean maximum skin score determined by the modified Rodnan method (2); both the right and the left sides are summed.

  • §

    Denominators and percentages reflect subjects objectively studied.

Vascular86 (99)303 (99)1.00
 Raynaud's phenomenon86 (99)300 (98)0.9636
  Digital pitting scars28 (32)162 (53)0.00098
  Digital tip ulcers21 (24)145 (47)0.00018
  Digital gangrene4 (5)54 (18)0.00428
 Amputations024 (8)0.0175
 Telangiectasias70 (80)256 (84)0.5392
Cutaneous   
 Total skin score, mean maximum4.15.50.030
  Finger skin thickness score, mean maximum3.03.60.016
 Puffy fingers73 (84)222 (72)0.0433
 Systemic sclerosis without scleroderma7 (8)20 (6.5)0.8020
Articular50 (57)175 (57)0.9626
 Palpable tendon friction rubs1 (1)2 (1)0.8188
Muscular5 (6)10 (3)0.3386
Gastrointestinal35/56 (62)§208/239 (87)0.0001
 Esophageal dysmotility23/45 (51)135/178 (76)0.0023
 Small bowel radiographic involvement8/27 (30)22/98 (22)0.6037
Pulmonary56/75 (74)130/256 (51)0.0004
 Pulmonary fibrosis33/68 (48)31/231 (13)<0.0001
 Restrictive lung disease15/56 (27)7/176 (4)<0.0001
 DLCO <65%37/56 (66)78/173 (45)0.0258
 Intrinsic PHT24 (28)59 (19)0.1936
 Pleuritis2 (2)3 (1)0.6700
Cardiac12/57 (21)34/213 (16)0.3639
Renal4 (5)00.0023
Other   
 Calcinosis13 (15)78 (25)0.0556
  Hand calcinosis (radiographic)6/21 (29)53/110 (48)0.1568
 Hypothyroidism10/84 (12)39/303 (13)0.9599
 Sicca findings4/31 (13)55/125 (44)0.0028

Skin thickening was minimal, as expected in lcSSc patients. The mean maximal total skin score was greater in the ACA group, primarily due to a higher regional skin thickness score for the fingers. In contrast, the anti-Th/To–positive patients more often had puffy fingers, a more subtle physical examination finding.

Gastrointestinal involvement was significantly more frequent in the ACA-positive patients, among those who were adequately studied, compared with the anti-Th/To–positive patients (87% and 62%, respectively), especially the occurrence of distal esophageal dysmotility (76% and 51%, respectively).

Pulmonary involvement overall was significantly more frequent in the anti-Th/To antibody group, among those who were adequately studied (74% versus 51% of ACA-positive patients). We used 2 measures of interstitial lung disease, radiographic evidence of pulmonary fibrosis and a restrictive pattern on pulmonary function testing; by both methods, interstitial lung disease was significantly more frequently detected in anti-Th/To–positive patients. Both groups had high frequencies of “intrinsic” pulmonary hypertension (anti-Th/To 28% and ACA 19%). Scleroderma renal crisis was uncommon but occurred exclusively in 4 patients (5%) with anti-Th/To antibodies. Sicca findings were significantly more frequently detected in the ACA group, but the proportion of patients adequately studied was low. There were no differences in articular, muscular, or cardiac abnormalities in the 2 patient groups.

Survival.

The anti-Th/To antibody group had significantly reduced cumulative survival from the time of the first Pittsburgh evaluation (P < 0.02) (Figure 1). The 5-year and 10-year cumulative survival rates for anti-Th/To–positive patients were 61% and 49%, respectively, whereas for the ACA-positive patients, these rates were 77% and 59%, respectively. “Intrinsic” pulmonary hypertension was the most common cause of death in both groups (Table 4). Pulmonary fibrosis was the next leading cause of death in anti-Th/To–positive patients, while cancer was the second most frequent known cause of death in the ACA group and was the fourth most frequent cause in the anti-Th/To group.

Figure 1.

Survival curves for the anti-Th/To antibody group compared with the anticentromere (ACA) antibody group with limited cutaneous systemic sclerosis, from the time of the first Pittsburgh evaluation. ∗ = P < 0.02, by Mantel-Haenszel test.

Table 4. Causes of death*
 Anti-Th/To (n = 32)ACA (n = 55)
  • *

    Values are the number (%) of patients. See Table 1 for definitions.

SSc-related23 (72)28 (51)
 Pulmonary hypertension14 (44)20 (36)
 Pulmonary fibrosis5 (15)2 (4)
 Cardiac2 (6)1 (2)
 Gastrointestinal2 (6)3 (5)
 Other02 (4)
Non–SSc-related9 (28)20 (36)
 Cancer4 (12)8 (14)
 Atherosclerosis02 (4)
 Other5 (15)10 (18)
Unknown07 (13)

DISCUSSION

Identification of the Th/To antigen was first reported in 1982 (12). Later studies further described Th/To as consisting of 2 RNA-processing enzymes, RNase MRP and RNase P, plus 10 associated proteins (13–17). Anti-Th/To antibody was initially shown to be a common antibody among patients with lcSSc (4). It is one of several autoantibodies detected in patients with SSc that produce nucleolar staining on routine ANA testing (5, 6, 8–10). Anti-Th/To antibody typically results in a bright nucleolar pattern or nucleolar and speckled pattern on staining (4).

Whether anti-Th/To antibody identifies a distinct subset of SSc is uncertain. To answer this question, we compared 2 large series of lcSSc patients who had either anti-Th/To antibodies or ACAs. Both groups were evaluated at the same institution during an identical time period and had an average disease duration since onset of >10 years at the time of their first evaluation (Table 2). A total of 800 patients with lcSSc was first evaluated at our institution during this time period. Anti-Th/To antibodies were present in 11.5% of these patients and ACAs in 41.4%. Thus, these 2 serum autoantibodies accounted for >50% of our lcSSc patients.

In many respects, the ACA and anti-Th/To groups were similar. More than 90% of the patients in each group had SSc as their clinical diagnosis, almost all having lcSSc. As expected in lcSSc, they had symptoms for a mean of 7–10 years prior to the first physician diagnosis of SSc. The frequency of articular, muscular, and cardiac involvement was similar. However, there were also differences, as noted in Tables 1–3. Anti-Th/To antibody–positive patients had more subtle skin changes, with a higher proportion having puffy fingers and a lower mean maximum total skin score, primarily due to lower scores for finger and hand skin thickening. Vascular disease was significantly less severe and sicca findings and involvement of the distal esophagus less frequent in anti-Th/To–positive patients. The diagnosis of SSc thus could easily be missed in patients with the anti-Th/To antibody.

“Intrinsic” pulmonary hypertension was present in 28% of anti-Th/To–positive patients and 19% of ACA-positive patients (P not significant). These proportions are considerably higher than those previously reported in the medical literature (18, 19). The percentages of anti-Th/To– and ACA–positive patients ultimately developing pulmonary hypertension among individuals who resided within 100 miles of Pittsburgh (our local referral area) at the time of first evaluation were 9 of 46 (20%) and 25 of 170 (15%), respectively. These proportions are similar to those in a recent report in which increased awareness of this complication was postulated as the reason for more frequent screening for “intrinsic” pulmonary hypertension (20). Thus, referral bias and increased surveillance may explain the increased frequency of pulmonary hypertension in our series.

We were surprised to find that patients with anti-Th/To developed 2 visceral complications more typically seen in dcSSc, namely scleroderma renal crisis and pulmonary fibrosis. Scleroderma renal crisis was uncommon (only 4 cases) but occurred exclusively in the anti-Th/To–positive patients. Interestingly, all 4 of these patients survived the scleroderma renal crisis but subsequently developed pulmonary hypertension that was fatal in 3 cases, as we have described in a previous report (21). Pulmonary fibrosis on chest radiograph was significantly more frequent in the anti-Th/To group (48% versus 13%; P < 0.0001). This proportion (48%) is similar to that typically reported in patients with anti–topoisomerase I antibodies (16), and its severity is attested to by the fact that pulmonary fibrosis was the second most common cause of death in the anti-Th/To–positive patients.

The mean duration between the first symptom of SSc and the first symptom of interstitial lung disease was significantly shorter in the anti-Th/To–positive patients (10.3 years) compared with the ACA-positive patients (15.7 years; P < 0.05). Thus, similar to the symptom-duration variables reported in Table 2, the anti-Th/To–positive patients had a temporal course that was somewhat more compressed than that of the ACA-positive patients. Among the male lcSSc patients, pulmonary fibrosis occurred significantly more frequently in those with anti-Th/To as compared with those with ACA (60% versus 7%; P = 0.0002). Of the 9 anti-Th/To–positive male lcSSc patients with pulmonary fibrosis, 5 were smokers and 2 had occupational silica dust exposure. There were only 2 ACA-positive male lcSSc patients with lung fibrosis, of whom 1 had both an environmental exposure to chemicals and a significant smoking history.

Although the anti-Th/To–positive patients with lcSSc were younger at first evaluation and came to our attention earlier in their disease, they had significantly worse survival. “Intrinsic” pulmonary hypertension was the most frequent cause of death in both groups. In contrast, Japanese SSc patients with ACA had a cumulative survival rate at 10 years after diagnosis of 96% (22). Causes of death in these Japanese patients were biliary cirrhosis and malignancy, and none developed pulmonary hypertension. Another report from Japan suggested that ACA in SSc patients might be considered a risk factor for the development of cancer (23). Malignancy was equally frequent in our ACA- and anti-Th/To antibody–positive patients (9.2% for each).

To emphasize the subtle nature of lcSSc in these patients, only 201 (66%) of the patients with ACA and 51 (59%) of the anti-Th/To–positive patients satisfied the American College of Rheumatology (ACR) classification criteria for definite SSc (24). It has previously been reported that these criteria do not adequately identify a sizable minority of lcSSc patients (25, 26). All of the individuals who did not satisfy the ACR criteria had other clinical evidence of SSc, including Raynaud's phenomenon in addition to 1 of the following: pulmonary arterial hypertension, pulmonary interstitial fibrosis, esophageal or small intestinal hypomotility, scleroderma heart disease, or inflammatory myopathy. The message for the family practitioner, internist, and rheumatologist is that all patients with Raynaud's phenomenon and a positive ANA should be carefully examined for an underlying systemic disorder, particularly SSc.

In summary, the presence of anti-Th/To antibodies identifies a subgroup of patients with lcSSc who have more subtle skin changes and less prominent vascular disease than that observed in lcSSc patients with ACA. The presence of the anti-Th/To antibody should be suspected in lcSSc patients whose ANAs show a bright nucleolar or nucleolar plus speckled staining pattern. Both anti-Th/To–positive and ACA-positive patients have a high frequency of “intrinsic” pulmonary hypertension, and anti-Th/To–positive patients also have an increased risk of pulmonary interstitial fibrosis and occasionally develop renal crisis. It should also be noted that lcSSc patients with anti-Th/To have reduced survival compared with ACA-positive lcSSc patients.

Ancillary