Interstitial Lung Disease in Anti–Jo-1 Patients With Antisynthetase Syndrome

Authors


Abstract

Objective

To assess the outcome of interstitial lung disease (ILD) in anti–Jo-1 patients with antisynthetase syndrome, determine predictive variables of ILD deterioration in these patients, and compare features of anti–Jo-1 patients with and without ILD.

Methods

Ninety-one anti–Jo-1 patients were identified by medical records search in 4 medical centers. All of these patients had undergone pulmonary function tests (PFTs) and high-resolution computed tomography (HRCT) scans.

Results

Sixty-six patients (72.5%) had ILD. Patients could be divided into 3 groups according to their presenting lung manifestations: acute onset of lung disease (n = 12), progressive onset of lung signs (n = 35), and asymptomatic patients exhibiting abnormalities consistent with ILD on PFTs and HRCT scans (n = 19). Sixteen patients had resolution of ILD; 39 and 11 patients experienced improvement and deterioration of ILD, respectively. ILD led to decreased functional status, since 29.8% of patients exhibited a marked reduction of activities due to ILD and 13.6% had respiratory insufficiency requiring oxygen therapy; 5 of 6 patients died due to ILD complications. Predictive parameters of ILD deterioration were HRCT scan pattern of usual interstitial pneumonia, respiratory muscle involvement, and age ≥55 years. Furthermore, anti–Jo-1 patients with ILD, compared with those without, more frequently exhibited mechanic's hands and lower creatine kinase levels.

Conclusion

Our findings confirm that ILD is a frequent complication in anti–Jo-1 patients, resulting in high morbidity. We suggest that patients with predictive factors of ILD deterioration may require more aggressive therapy. Finally, anti–Jo-1 patients with ILD, compared with those without, may exhibit a particular clinical phenotype.

INTRODUCTION

Antisynthetase syndrome is a clinical condition characterized by polymyositis/dermatomyositis (PM/DM) associated with antisynthetase antibodies, arthritis, Raynaud's phenomenon, mechanic's hands, and interstitial lung disease (ILD) (1). To date, 8 antisynthetase antibodies have been identified (2), with anti–Jo-1 antibody being the most common (1–5). Among clinical extramuscular manifestations, ILD is the most frequent complication of antisynthetase syndrome (70–89%) (2–12), resulting in increased morbidity in these patients (12–16).

However, to date, only a few authors have accurately evaluated the outcome of ILD in anti–Jo-1–positive patients (10, 12, 17). The aims of the current study were to assess the outcome of ILD in anti–Jo-1–positive patients and determine predictive variables of ILD deterioration in these patients. We further compared features of anti–Jo-1–positive patients with and without ILD to determine whether anti–Jo-1 patients with ILD exhibit a particular clinical phenotype.

Significance & Innovations

  • To date, only a few authors have evaluated the outcome of interstitial lung disease (ILD) in anti-Jo-1 patients with antisynthetase syndrome.

  • Our series underscores that ILD results in high morbidity in anti–Jo-1 patients. Therefore, 16.7% of patients experienced deterioration of ILD; ILD further resulted in decreased functional status, since 29.8% of patients exhibited a marked reduction of activities due to ILD and 13.6% had respiratory insufficiency requiring oxygen therapy.

  • In anti–Jo-1 patients, predictive parameters of ILD deterioration were high-resolution computed tomography scan pattern of usual interstitial pneumonia, respiratory muscle involvement, and age ≥55 years.

  • Anti–Jo-1–positive patients with ILD, compared with those without, may exhibit a particular phenotype characterized by more frequent mechanic's hands and lower values of blood creatinine kinase.

PATIENTS AND METHODS

This retrospective study began with a search of the institutional centers' medical record index, which provides access to the diagnoses of the centers' patients. The first electronic search involved use of the codes PM/DM to identify patients with a diagnosis of PM/DM who were seen as either inpatients or outpatients in 4 academic centers (Lille, Cochin, Pitié-Salpêtrière, and Rouen) between 1996 and 2010. The diagnosis of PM/DM was based on criteria by Bohan and Peter (18, 19). During the study period, 346 consecutive patients were seen for evaluation of PM/DM; no patient had amyopathic DM. All patients had been tested for anti–Jo-1 antibody using immunodiffusion with subsequent confirmation by enzyme-linked immunosorbent assay. A second search was used to isolate the subset of anti–Jo-1–positive patients; 91 anti–Jo-1–positive patients were identified. All of these patients had routinely undergone lung investigations to depict ILD, i.e., pulmonary function tests (PFTs) and pulmonary high-resolution computed tomography (HRCT) scans. Therefore, a third search was made to isolate the subset of anti–Jo-1–positive patients with ILD; 66 anti–Jo-1–positive patients with antisynthetase syndrome had ILD (72.5%). Ethical approval was obtained from the local committee (CERNI for the CPP de Haute-Normandie).

All 66 anti–Jo-1–positive patients had an initial evaluation of organ involvement that resulted in the detection of systemic complications, including 1) Raynaud's phenomenon; 2) mechanic's hands; 3) joint impairment; 4) esophageal dysfunction (i.e., dysphagia, gastroesophageal reflux, coughing while eating, and aphagia for solids/liquids); 5) respiratory muscle involvement resulting from ventilatory insufficiency related to striated muscle weakness when patients had ventilatory failure with decreased vital capacity (VC); ventilatory insufficiency due to respiratory striated muscle weakness was dichotomized as previously described (14): severe hypoventilation, determined by hypercapnic respiratory failure requiring mechanical ventilation, and moderate hypoventilation, characterized by a restrictive pattern on PFTs (decreased lung volumes with VC <80%) without evidence of ILD; and 6) aspiration pneumonia. Patients were also examined for underlying malignancy. Moreover, patients underwent a biochemical assessment using creatine kinase (CK).

Evaluation of ILD.

In all 66 anti–Jo-1–positive patients, the initial diagnosis of ILD was based on appropriate clinical data, PFTs, and HRCT scan abnormalities.

Pulmonary symptoms.

The clinical symptoms assessed were dyspnea, cough, and fever. Patients were also classified into 3 groups according to the initial clinical presentation of ILD, as previously described (12, 17, 20): acute onset of ILD, symptomatic progressive onset of lung symptoms, and asymptomatic disorder with abnormalities consistent with ILD on PFTs and HRCT scans.

PFTs.

Assessments of pulmonary function were made according to the American Thoracic Society (ATS) guidelines using standard equipment. The following parameters were evaluated at ILD diagnosis: VC, forced VC (FVC), and diffusing capacity for carbon monoxide (DLCO). VC and FVC were measured by spirometry (using a water-sealed spirometer); DLCO was obtained by the single-breath method. Data are expressed as percentages of the predicted values. The predicted values for each subject, based on sex, age, height, and weight, were obtained from standard tables (21). Lung function was considered abnormal when volumes were <80% of the predicted values and when DLCO was <70% of the predicted values.

Pulmonary HRCT scan.

At ILD diagnosis, an HRCT scan was performed to evaluate abnormalities consistent with ILD. The diagnostic criteria for ILD were based on the ATS classification of idiopathic ILD, i.e., the presence of (22) parenchymal micronodules/nodules, linear opacities (septal lines), irregularity of the interfaces between peripheral pleura and aerated lung parenchyma, ground-glass opacities, consolidation, honeycombing, and subpleural cysts/traction bronchiectases or bronchiolectases (22).

The severity of ILD on HRCT scans was scored according to Warrick et al (23). A point value was assigned to each abnormality: ground-glass opacities (point value 1), irregular pleural margins (point value 2), septal/subpleural lines (point value 3), honeycombing (point value 4), and subpleural cysts/traction bronchiectases/bronchiolectases (point value 5). A severity of ILD score was generated by adding point values for HRCT abnormalities, with the severity score having a possible range of 0 (normal) to 15 (all abnormalities presented). An extent of ILD score was generated by localizing each abnormality to the number of bronchopulmonary segments involved, as follows: 1–3 segments (point value 1), 4–9 segments (point value 2), and >9 segments (point value 3). This process was repeated for each of the abnormalities, allowing the possible extent of ILD scores of 0 (normal) to 19 (>9 segments showing all 5 abnormalities). The severity and extent of ILD scores were then added to form a total HRCT score, with a possible range of 0–30 (23).

HRCT pattern has been correlated with pulmonary histologic findings, including 1) cryptogenic organizing pneumonia (COP), characterized by consolidation and linear opacities; 2) nonspecific interstitial pneumonia (NSIP), characterized by ground-glass opacities and irregular linear opacities; 3) usual interstitial pneumonia (UIP), characterized by honeycombing and traction bronchiectases; and 4) diffuse alveolar damage (DAD), defined by bilateral and extensive consolidation with airspace and ground-glass opacities (6, 24–28). Our patients were therefore divided into 4 groups based on the predominant pattern on HRCT: COP, NSIP, UIP, or DAD, as previously described (6, 24–28). In our patients, both the severity and classification of HRCT abnormalities were based on independent analyses of an expert chest radiologist (AJ) and pneumologist (SD), both of whom were blinded to clinical features; discrepant findings were re-reviewed by both experts to determine consensus regarding severity and classification of ILD.

Lung biopsy.

Twenty-two patients underwent transbronchial or surgical lung biopsy at ILD diagnosis. Histologic analysis of lung biopsy specimens was performed to detect abnormalities consistent with ILD: COP, NSIP, UIP, or DAD (26).

Outcome of ILD.

The outcome data for ILD were assessed using validated tools (lung symptoms, PFTs, and HRCT findings) at the last date of followup. The following treatment regimens and changes in ILD therapy during followup were recorded: 1) steroids (at an initial dosage of 1 mg/kg/day), 2) azathioprine (2 mg/kg/day), 3) cyclophosphamide (6 pulses at 0.7 gm/m2/month), and 4) mycophenolate mofetil (30 mg/kg/day). The response to treatment of ILD was evaluated according to the ATS criteria (29). The outcome of ILD was categorized as resolution, improvement, or deterioration. Resolution was defined as complete resolution of pulmonary symptoms associated with the disappearance of radiographic signs of ILD and normalization of standard PFT values. Improvement was defined as when any of the former pulmonary alterations improved without returning to the normal value; according to an international consensus statement of the ATS on idiopathic ILD (29), changes of ≥10% in FVC and/or ≥15% in DLCO were considered to be significant, and were used as determinants of improvement. Deterioration was defined as when any of the former pulmonary conditions worsened despite institution of therapy; changes of ≥10% in FVC and/or ≥15% in DLCO were considered to be significant, and were used as determinants of deterioration (29). Survival status and cause of death were based on physician records.

Statistical analyses.

First, we compared the features of antisynthetase syndrome patients with and without ILD deterioration. Second, we compared antisynthetase syndrome–related manifestations between anti–Jo-1–positive patients with and without ILD. For group comparisons involving binary data, we used either the chi-square test or Fisher's exact test, depending on the sample size (n = >5 and n = ≤5, respectively). Comparisons involving continuous data were made using the Mann-Whitney U test. The results were regarded as significant when the P value was less than 0.05. With regard to variables with a P value less than 0.1, we further proceeded with multiple logistic regression to calculate the multivariate odds ratio (OR) and 95% confidence interval (95% CI); the level of significance used was P values less than 0.05 in all performed tests.

RESULTS

The 66 anti–Jo-1–positive patients comprised 25 men and 41 women with a median age of 55 years (range 25–74 years) at ILD diagnosis. Forty-eight patients had PM and 18 had DM. The patients presented antisynthetase syndrome–related complications, including Raynaud's phenomenon (n = 32), mechanic's hands (n = 23), joint impairment (n = 44), esophageal dysfunction (n = 11), ventilatory insufficiency due to striated muscle weakness (n = 7), and aspiration pneumonia (n = 6); 3 patients had cancer.

Characteristics of ILD.

ILD onset preceded initial PM/DM clinical manifestations in 10 patients; the median interval onset between ILD diagnosis and PM/DM was 6 months. ILD was concurrently identified in association with PM/DM in 42 patients. ILD developed after PM/DM onset in 14 patients; the median length of time before onset of ILD after the diagnosis of PM/DM was 18 months.

At ILD diagnosis, pulmonary symptoms consisted of 1) dyspnea (n = 44), which was classified, according to the New York Heart Association classification, as stages II (n = 32), III (n = 10), and IV (n = 2); 2) cough (n = 28); and 3) fever (n = 16). Patients were divided into 3 groups according to their presenting lung manifestations: acute onset of ILD (n = 12), symptomatic progressive onset of ILD (n = 35), and asymptomatic patients exhibiting abnormalities consistent with ILD on PFTs and HRCT scans (n = 19).

At ILD diagnosis, PFT findings were consistent with ILD in all cases (Table 1). At ILD diagnosis, HRCT scans of the lungs also demonstrated abnormalities consistent with ILD, including parenchymal micronodules or nodules (n = 19), linear opacities (n = 49), irregularity of the interfaces (n = 57), ground-glass opacities (n = 53), honeycombing (n = 13), consolidation (n = 11), and traction bronchiectases/bronchiolectases (n = 9). The median value of the fibrosis score on HRCT was 14.9 at ILD diagnosis. Furthermore, according to the predominant pattern on the pulmonary HRCT scans, patients could be divided into 3 groups: COP (n = 11), NSIP (n = 39), and UIP (n = 16).

Table 1. ILD characteristics of 66 anti–Jo-1–positive patients with antisynthetase syndrome*
 Antisynthetase syndrome with ILD
  • *

    Values are the number (percentage) unless otherwise indicated. ILD = interstitial lung disease; PM/DM = polymyositis/dermatomyositis; PFT = pulmonary function test; FVC = forced vital capacity; VC = vital capacity; DLCO = diffusing capacity for carbon monoxide; HRCT = high-resolution computed tomography; COP = cryptogenic organizing pneumonia; NSIP = nonspecific interstitial pneumonia; UIP = usual interstitial pneumonia.

Presenting symptoms 
 Symptomatic acute onset of lung disease12 (18.2)
 Symptomatic progressive onset of lung signs35 (53)
 Asymptomatic19 (28.8)
Time of onset 
 Before PM/DM10 (15.2)
 Concomitant with PM/DM42 (63.6)
 After PM/DM14 (21.2)
PFT findings at ILD diagnosis, % 
 FVC73
 VC74
 DLCO59
HRCT scan pattern 
 COP11 (16.7)
 NSIP39 (59.1)
 UIP16 (24.2)
Course 
 Resolution16 (24.2)
 Improvement39 (59.1)
 Deterioration11 (16.7)
Mortality6 (9.1)

Pulmonary biopsy specimens from 22 patients were subjected to histologic analysis, which showed damage consistent with COP (n = 6), NSIP (n = 11), and UIP (n = 5). In all 22 patients, we found a correlation between the predominant pattern on HRCT scans and histologic pulmonary damage. Therefore, the 6 patients with histologic findings of COP exhibited consolidation/linear opacities on HRCT scans. The 11 patients with histologic findings of NSIP had ground-glass opacities and irregular linear opacities on HRCT scans, whereas the 5 patients with histologic UIP had honeycombing and traction bronchiectases on HRCT scans.

Course of ILD.

None of the 66 patients were lost to followup.

ILD resolution.

ILD resolved in 16 patients (24.2%), i.e., lung symptoms completely healed and PFTs returned to normal associated with clearing of HRCT scans (Table 1). Two of these patients exhibited acute onset of ILD; the other patients had symptomatic progressive onset of ILD (n = 10) or the asymptomatic form at ILD diagnosis (n = 4). The median interval between the onset of ILD symptoms and therapy institution was 11 months. Complete control of ILD was obtained with 1) steroids alone (n = 12), 2) steroids and azathioprine (n = 2), 3) steroids and cyclophosphamide (n = 1), and 4) steroids/azathioprine, and mycophenolate mofetil subsequently because of azathioprine inefficacy (n = 1). None of these patients had died at the last followup.

ILD improvement.

Thirty-nine PM/DM patients (59.1%) experienced ILD improvement (n = 28) or stabilization (n = 11). Lung symptoms disappeared in 16 patients, who concomitantly exhibited a decrease of abnormalities on PFTs and HRCT scans. The 23 other patients had improvement in clinical pulmonary signs, and results of the PFTs and HRCT scans also improved without reaching normal patterns. The 39 patients exhibited 1) acute onset of ILD (n = 6), 2) symptomatic progressive onset of ILD (n = 19), or 3) the asymptomatic form (n = 14). The median interval between the onset of ILD symptoms and therapy initiation was 8 months. Patients received the following therapy for ILD: 1) steroids alone (n = 16), 2) steroids and azathioprine (n = 9), 3) steroids and cyclophosphamide (n = 6), 4) steroids/cyclophosphamide and azathioprine subsequently (n = 4), and 5) steroids/azathioprine, and mycophenolate mofetil subsequently because of azathioprine inefficacy (n = 4). In the 5 patients with oxygen dependency at the initial diagnosis of ILD, oxygen therapy could be stopped in 4 cases under treatment. One of the 39 patients died of myocardial infarction at the 52-month followup.

ILD deterioration.

In 11 patients (16.7%), pulmonary symptoms worsened despite therapy; ILD deterioration was concurrently reflected on PFTs and HRCT scans. Four of these patients had an acute onset of ILD; the 7 other patients exhibited symptomatic progressive onset of ILD (n = 6) or the asymptomatic form (n = 1). The median interval between the onset of ILD symptoms and the beginning of therapy was 6 months. Patients received the following therapy for ILD: 1) steroids alone (n = 3), 2) steroids and azathioprine (n = 2), 3) steroids and cyclophosphamide (n = 1), 4) steroids/cyclophosphamide and azathioprine subsequently (n = 4), and 5) steroids/cyclophosphamide and azathioprine subsequently, azathioprine being inefficient, it was replaced by mycophenolate mofetil (n = 1). Finally, 7 patients developed progressive respiratory failure within a median delay of 23.7 months, resulting in oxygen dependency despite therapy. Five patients (7.6%) died within a median interval of 20 months after ILD diagnosis; death was due to pneumonia (n = 3) and respiratory failure (n = 2).

Factors associated with ILD deterioration.

The median time from ILD diagnosis and therapy initiation did not differ between patients with ILD resolution/improvement and those who exhibited ILD deterioration (2 months versus 1.5 months; P = 0.842). Furthermore, the median duration of ILD followup was 36 months (range 5– 163 months) for the overall population; it was not different between the patients who experienced resolution/improvement or deterioration of ILD (P = 0.736).

Anti–Jo-1 patients with ILD deterioration were older (at ILD diagnosis) than those without ILD (63 years versus 53.5 years) (Table 2). Moreover, patients with ILD deterioration more commonly had ventilatory insufficiency due to striated muscle weakness (P = 0.01) and aspiration pneumonia (P = 0.05) (Table 2).

Table 2. Comparison of clinical characteristics between anti–Jo-1 antisynthetase syndrome patients with and without ILD deterioration*
 ILD deterioration (n = 11)Absence of ILD deterioration (n = 55)P
  • *

    ILD = interstitial lung disease; PM/DM = polymyositis/dermatomyositis; ESR = erythrocyte sedimentation rate.

  • Obtained with chi-square or Fisher's exact tests.

General characteristics   
 Age, median (range) years63 (47–73)53.5 (25–74)0.007
 Sex, %  0.184
  Male18.241.8 
  Female81.858.2 
 PM/DM subset, %81.8/18.270.9/29.10.713
Clinical characteristics, %   
 Raynaud's phenomenon54.534.80.515
 Mechanic's hands54.536.40.319
 Esophageal involvement9.110.91
 Joint involvement63.665.51
 Ventilatory insufficiency36.45.50.01
 Aspiration pneumonia36.410.90.05
 Cancer05.51
Biochemical parameters, median (range)   
 ESR, mm/hour28 (15–745)19 (2–108)0.476
 Creatine kinase, IU/liter225 (50–4,601)297 (33–20,000)0.521

The group of patients with an acute onset of ILD tended to have ILD deterioration more frequently, although not significantly so (P = 0.103). At the initial ILD diagnosis, patients who exhibited ILD deterioration had lower median values of DLCO (P = 0.002) (Table 3). With regard to HRCT findings, the median score of fibrosis (at ILD diagnosis) did not differ between patients with and without ILD deterioration (P = 0.385). Only the UIP pattern on HRCT scans was more frequent in the group of patients with ILD deterioration (P = 0.01). At the last PFT followup, patients with ILD deterioration presented lower median values of FVC (P < 10−5), VC (P < 10−5), and DLCO (P < 10−6), and a higher median score of fibrosis (P = 0.004). With regard to ILD therapy, steroid-refractory ILD tended to be found more frequently in patients experiencing ILD deterioration (72.7% versus 49.1%). The mortality rate was higher in patients with ILD deterioration compared with those without (36.4% versus 1.8%; P = 0.002).

Table 3. Comparison of pulmonary characteristics between anti–Jo-1 antisynthetase syndrome patients with and without ILD deterioration*
 ILD deterioration (n = 11)Absence of ILD deterioration (n = 55)P
  • *

    Values are the percentage. ILD = interstitial lung disease; PM/DM = polymyositis/dermatomyositis; PFT = pulmonary function test; FVC = forced vital capacity; VC = vital capacity; DLCO = diffusing capacity for carbon monoxide; HRCT = high-resolution computed tomography; COP = cryptogenic organizing pneumonia; NSIP = nonspecific interstitial pneumonia; UIP = usual interstitial pneumonia.

Presenting pulmonary symptoms   
 Fever18.225.51
 Dyspnea90.961.80.08
 Cough63.638.20.182
Initial ILD presentation  0.311
 Acute onset of ILD36.414.5 
 Progressive onset of ILD54.560 
 Asymptomatic form9.125.5 
Time of onset  0.706
 Before PM/DM18.214.5 
 Concomitant with PM/DM54.565.5 
 After PM/DM27.320 
PFT findings at ILD diagnosis   
 FVC6875.50.846
 VC70750.938
 DLCO47600.002
HRCT scan pattern at ILD diagnosis   
 HRCT score of fibrosis, median16.212.90.385
 COP0200.186
 NSIP45.561.80.336
 UIP54.518.20.01
Mortality36.41.80.002

Under multivariate analysis, significant factors for ILD deterioration in anti–Jo-1 patients were UIP pattern on HRCT scans of the lungs (OR 5.02 [95% CI 1.05–243], P = 0.04), ventilatory insufficiency due to striated muscle weakness (OR 10.95 [95% CI 1.46–82.3], P = 0.02), and age ≥55 years (OR 5.53 [95% CI 0.84–36.6], P = 0.07), with the age cutoff ≥55 years being used because the median age of our patients was 55 years.

Comparison of anti–Jo-1–positive patients with and without ILD.

The median duration of followup did not differ between anti–Jo-1 patients with and without ILD deterioration (P = 0.862). We found no statistically significant differences between anti–Jo-1–positive patients with and without ILD for age, sex, PM/DM subset, myalgia (P = 0.247), muscle weakness (P = 0.472), Raynaud's phenomenon, esophageal impairment, ventilatory insufficiency related to striated muscle weakness (P = 1), and cancer (Table 4). However, anti–Jo-1–positive patients with ILD more commonly exhibited mechanic's hands (P = 0.009) and lower CK levels at diagnosis of antisynthetase syndrome (P = 0.02).

Table 4. Comparison of clinical characteristics between anti–Jo-1 antisynthetase syndrome patients with and without ILD*
 With ILD (n = 66)Without ILD (n = 25)P
  • *

    ILD = interstitial lung disease; PM/DM = polymyositis/dermatomyositis.

  • Obtained with chi-square or Fisher's exact tests.

General characteristics   
 Age, median (range) years55 (25–74)57 (18–79)0.586
 Sex, %  1
  Male37.936 
  Female62.164 
 PM/DM subset, %72.7/27.352/480.08
Clinical characteristics, %   
 Raynaud's phenomenon48.5400.491
 Mechanic's hands34.880.009
 Esophageal involvement16.7360.08
 Joint involvement66.7600.626
 Ventilatory insufficiency10.6121
 Aspiration pneumonia9.140.668
 Cancer4.5160.08
Biochemical parameter, median (range)   
 Creatine kinase, IU/liter273 (50–8,109)500 (24–20,000)0.02

Antisynthetase syndrome therapy was not different in anti–Jo-1 patients with and without ILD antibodies for methotrexate (24.2% versus 28%; P = 0.788), azathioprine (40.9% versus 20%; P = 0.08), intravenous immunoglobulins (31.8% versus 40%; P = 1), and mycophenolate mofetil (9.1% versus 0%; P = 0.182); however, anti–Jo-1 patients with ILD more often received cyclophosphamide (25.8% versus 4%; P = 0.01).

Finally, the mortality rate did not differ between anti–Jo-1 patients with and without ILD (9.1% versus 16%; P = 0.453); however, the causes of death were not similar between the 2 groups of patients. Anti–Jo-1 patients without ILD died of cancer (n = 3 of 4) and heart failure unrelated to antisynthetase syndrome (n = 1 of 4). Five (83.3%) of 6 patients with ILD died of pulmonary complications; the sixth patient died of myocardial infarction unrelated to antisynthetase syndrome.

DISCUSSION

ILD occurs in 70–90% of anti–Jo-1 patients with antisynthetase syndrome (1–12, 17, 25, 30). However, to date, the outcome of ILD still remains poorly defined in these patients. Although the characteristics of our anti–Jo-1 patients with ILD may be impacted by referral bias, our findings confirm the high frequency of ILD (72.5%) in such patients. To the best of our knowledge, our study is the largest cohort to assess the outcome of ILD in anti–Jo-1 patients with antisynthetase syndrome. Because our population of patients (who were referred to tertiary care centers) was reasonably homogeneous, we believe that the evaluation and response to treatment of ILD, under standard conditions, is possible.

ILD time of onset is variable in antisynthetase syndrome (31, 32). Older series have described that myositis preceded ILD (7–33%), was concurrently identified with ILD (38–60%), and occurred after ILD (29–50%) (9, 33). Our anti–Jo-1 patients with ILD had a similar composition. Our findings reinforce that the detection of anti–Jo-1 antibody is useful in patients with ILD for predicting underlying antisynthetase syndrome. Further, screening for subclinical ILD should be made routinely in anti–Jo-1 patients, resulting in both diagnosis of and therapy for ILD at an earlier stage.

Patients with antisynthetase syndrome have been reported to exhibit chronic and mild ILD (9, 34). In 10 antisynthetase syndrome patients, the initial presentation of ILD was 20%, 30%, and 50% in the acute, progressive, and asymptomatic forms, respectively (9). In 32 antisynthetase syndrome patients, the frequency of acute onset of ILD was reported to be higher (47%); however, these patients were followed up in departments of pneumology, which may result in a bias of recruitment (severe forms of ILD). Most of our anti–Jo-1 patients developed a nonacute onset of ILD (82%); therefore, ILD should be viewed as a clinical spectrum in antisynthetase syndrome, ranging from acute pulmonary presentation to symptomatic progressive or asymptomatic ILD. One previous pneumologic series has found that acute onset of ILD was a poor prognostic factor in antisynthetase syndrome (12); by contrast, we failed to find a correlation between ILD deterioration and acute onset of ILD in our patients. Interestingly, we observed that respiratory muscle involvement was associated with ILD deterioration; these patients with concurrent ILD deterioration might exhibit a more severe form of antisynthetase syndrome.

Our study shows that PFTs and HRCT scans may provide prognostic information in anti–Jo-1 patients with ILD. Reduced FVC and DLCO values have been found to be poor prognostic factors in PM/DM patients with ILD (35–37). We have also observed a correlation between ILD deterioration and lower values of DLCO at ILD diagnosis. Furthermore, HRCT scan findings have been described in antisynthetase syndrome, with linear and ground-glass opacities being the most frequent signs (10, 12, 17, 25). In our patients, linear opacities, irregularity of the interfaces, and ground-glass opacities were also the most common HRCT scan abnormalities (78.3–86.7%). Another interesting finding of the present study was that HRCT scans of the lungs may provide data regarding ILD pattern, as suggested previously (26–28). Verschakelen (38) has recently underscored that it is possible to correctly diagnose ILD by HRCT scan when typical patterns of ILD are present, so that lung biopsy may be avoided in most cases. In the current study, not all patients underwent lung biopsy because this test is an invasive method, and therefore it is not possible to conclude definitively that the HRCT scan pattern is correlated with histologic data. However, among our 22 anti–Jo-1 patients who underwent a lung biopsy, we found a correlation between HRCT scan pattern and histologic damage in all cases. Our patients with antisynthetase syndrome more commonly experienced an NSIP pattern (59.1%) on HRCT scans; these findings are in accordance with previous data (25). In contrast, in a series of antisynthetase syndrome patients undergoing lung biopsy, a preponderance of UIP (n = 10) and DAD (n = 12) was described (17); the authors noted that the predominance of UIP/DAD may reflect procedural bias (sicker patients requiring biopsy are more likely to have severe ILD, i.e., DAD and UIP). We have interestingly shown a marked correlation between histologic damage of UIP and the onset of ILD deterioration, which suggests that these patients may require more aggressive therapy.

To date, only a few investigators have assessed the outcome of ILD in patients with antisynthetase syndrome (1, 9, 10). Previous HRCT series in which ILD was followed up have found an improvement in 6.7%, stabilization in 46.7%, and a worsening in 15% of patients with antisynthetase syndrome (10, 25, 31). In 32 antisynthetase syndrome patients with ILD, the majority of patients (72%) stabilized with therapy; however, 28% developed respiratory failure and died. Regarding the 1-year outcome of ILD, no difference was found between patients with acute and nonacute onset of ILD (12). More recently, in 13 antisynthetase syndrome patients with ILD, the patients exhibited worsening (60%) and stabilization (40%) of ILD at the 1-year followup (25). In our larger cohort of anti–Jo-1 patients, ILD was associated with high morbidity. Therefore, we have found that only 16 patients (24.2%) had resolution of ILD, and 39 patients (59.1%) and 11 patients (16.7%) exhibited improvement and deterioration of ILD, respectively. In these latter patients, we have shown that ILD resulted in a marked decrease of functional status. At the last followup, 28.8% of our overall patients exhibited a marked reduction of activities due to ILD and 13.6% had respiratory insufficiency requiring oxygen therapy. Finally, 7.6% of the overall anti–Jo-1 patients with ILD died of lung complications.

Although the optimal therapy of ILD has not yet been clearly established, corticosteroid therapy is considered the first-line therapy for PM/DM patients with ILD (14, 38, 39). The therapeutic response to steroids mainly depends on pulmonary histologic findings; COP and NSIP are considered to produce the better steroid response forms in PM/DM (14, 39). The present study also demonstrates an increased response to steroid therapy in antisynthetase syndrome patients with NSIP and COP compared with those with UIP, which suggests that early control of alveolitis may be required before it causes irreversible lung damage. Moreover, patients with steroid-refractory ILD tended to develop deterioration more commonly than those without. Cyclophosphamide may improve the clinical outcome in these patients; intravenous cyclophosphamide is most often used, although an oral regimen might be preferred by some investigators (14, 26, 39–42). In a retrospective series, cyclophosphamide (0.5–0.8 gm/m2 every 4 weeks for 6 cycles) was used in 17 PM/DM patients with ILD; at the 7-month followup, dyspnea improved in 67% of patients and VC improved significantly in 47% of cases (41). Our retrospective series also suggests that pulse cyclophosphamide is useful in anti–Jo-1 patients with ILD. In this instance, 18 anti–Jo-1–positive patients with ILD were given intravenous pulse cyclophosphamide (0.7 gm/m2/month for 6 cycles); we have interestingly observed that cyclophosphamide resulted in resolution or significant improvement of pulmonary status in 16.7% and 55.6% of the overall cyclophosphamide-treated patients, respectively. Only 1 case report of azathioprine as a helpful adjunctive maintenance therapy for the control of ILD has been described in PM/DM (43). In our experience, azathioprine was the most commonly used steroid-sparing agent in the therapy of antisynthetase syndrome–related ILD. In the 27 azathioprine-treated patients with ILD, azathioprine was given either after the administration of cyclophosphamide as a maintenance therapy (33.3%) or in combination with prednisone (66.7%); in these 27 patients, resolution or significant improvement of the pulmonary status was observed in 10.5% and 47.4% of cases, respectively. Furthermore, an open trial included 5 patients with ILD-associated PM/DM who received mycophenolate mofetil; at the 18-month followup, the PFTs did not improve significantly (44). On the other hand, in a small series, 3 PM/DM patients with ILD received steroids and mycophenolate mofetil, which resulted in resolution of ILD (45). In our experience, 6 anti–Jo-1–positive patients with refractory ILD received mycophenolate mofetil, which resulted in improvement of ILD (83.3%). Taken together, our findings suggest that mycophenolate mofetil may be useful in anti–Jo-1 patients with ILD.

Finally, from a practical point of view, the knowledge of predictive factors of ILD onset appears essential to improve the management of anti–Jo-1 patients. Indeed, an interesting finding in the present series was that, among antisynthetase syndrome–related extrapulmonary manifestations, only hyperkeratotic-rhagadiform hand symptoms (mechanic's hands) were more frequently encountered in the group of anti–Jo-1–positive patients with ILD. On the other hand, we did not observe a relationship between cancer and ILD because cancer prevalence was similar between anti–Jo-1 patients with and without ILD; therefore, our data do not corroborate the hypothesis of a paraneoplastic form of ILD in anti–Jo-1 patients with antisynthetase syndrome. In addition, Fudman and Schnitzer (46) reported that PM/DM patients with normal CK levels are at risk of ILD; other authors further described that muscle lesions may be absent in PM/DM patients with Hamman-Rich syndrome (47). The present study emphasizes that the median value of serum CK was lower in the group of anti–Jo-1–positive patients with ILD. Our latter data, in fact, suggest that muscle damage differs from ILD in anti–Jo-1 patients with antisynthetase syndrome. However, further investigations are warranted to confirm our data.

In conclusion, our series highlights that ILD results in increased morbidity in anti–Jo-1 patients. Our study also suggests that the following parameters could be considered as predictive of a poor outcome of ILD: older age; symptomatic ILD; lower values of FVC, VC, and DLCO at ILD diagnosis; UIP pattern; and steroid-refractory ILD. The presence of these factors may suggest a more aggressive therapy of anti–Jo-1 patients with ILD. Finally, our findings suggest that anti–Jo-1–positive patients with ILD, compared with those without, may exhibit a particular phenotype characterized by more frequent mechanic's hands and lower values of blood CK.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Marie had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Marie, Josse, Dominique, Jouen.

Acquisition of data. Marie, Josse, Hatron, Dominique, Hachulla, Janvresse, Cherin, Mouthon, Vittecoq, Menard, Jouen.

Analysis and interpretation of data. Marie, Josse, Dominique, Menard.

Ancillary