Pulmonary Pathologic Manifestations of Anti-Alanyl-tRNA Synthetase (Anti–PL-12)–Related Inflammatory Myopathy

The University of Pittsburgh's Idiopathic Inflammatory Myopathy registry was searched for patients who, at the time of their initial clinical evaluation, were found to have anti–PL-12 antibodies. Anti–PL-12 was identified by radiolabeled protein immunoprecipitation, with positive results confirmed by RNA immunoprecipitation, as previously described.⁷ 

Clinical features were extracted from this myositis registry, which encompasses more than 3 decades of prospective data and serum collected on consecutive outpatients and inpatients with myositis evaluated at the University of Pittsburgh (Pittsburgh, Pennsylvania). The extracted database was complemented by retrospective review of the electronic medical record. Patients possessing anti–PL-12 autoantibodies who were initially seen between January 1985 and December 2012 were included, regardless of their connective tissue disease (CTD) diagnosis. Patients with myositis met the diagnostic criteria of Bohan and Peter,⁸  whereas the diagnoses of systemic sclerosis, Sjögren syndrome, undifferentiated CTD (UCTD), or other overlapping rheumatic syndromes were made clinically by experienced rheumatologists.

All variables (clinical, laboratory, radiographic, and pathologic) and organ system definitions were well defined and standardized in this registry. Organ involvement was defined as follows: (1) vascular, the presence of Raynaud phenomenon, digital pitting scars, ulcers and gangrene, or abnormal nailfold capillaries; (2) cutaneous, characteristic rashes of DM or sclerodactyly; (3) joints, objective joint swelling and tenderness; (4) muscle, objective, proximal muscle weakness, in addition to any one of the following: elevated serum creatine kinase, myopathic electromyogram, or myositis on muscle biopsy; (5) gastrointestinal, proximal or distal, esophageal dysmotility or small/large bowel involvement; (6) pulmonary, fibrosis on chest radiograph or high-resolution chest computed tomography; and (7) primary pulmonary artery hypertension, mean pulmonary artery pressure more than 25 mm Hg on cardiac catheterization or systolic pressure more than 40 mm Hg estimated by echocardiography in the absence of significant pulmonary and/or cardiac disease causing secondary pulmonary artery hypertension. Interpretations of chest computed tomographies were performed by thoracic radiologists at our institution.

Only patients for whom histology slides of lung tissue were available for review were included in the study. An average number of 12 slides stained with hematoxylin-eosin were available for study (median, 7; range, 2–52). The histopathologic pattern of lung involvement was classified according to the American Thoracic Society/European Respiratory Society consensus classification of idiopathic interstitial pneumonias.^9,10  A clinical diagnosis of ILD required restrictive physiology on pulmonary function testing and abnormal radiologic findings by high-resolution chest computed tomography characterized by at least one of the following: reticulation, traction bronchiectasis, honeycomb change, or ground-glass opacities.

Specimens from 12 patients were available for study; 10 patients had surgical lung biopsies (1 underwent bilateral lung transplantation 3 years later; 1 underwent transbronchial biopsy 5 years later; both were also available for review), and 2 patients underwent single and double lung transplantation, respectively.

The clinical features of the 12 patients with anti–PL-12 autoantibodies are shown in Table 1. The patients' median age was 52 years (range, 17–64), with 9 woman and 3 men. Most patients presented with dyspnea (9 of 75; 67%) as their initial symptom. Four patients carried a diagnosis of PM, 5 had UCTD, 2 had systemic sclerosis, and 1 had Sjögren syndrome. The most common ARS features, other than dyspnea, were erythematous skin rashes in 5 of 12 patients (42%), sclerodactyly in 2 of 12 patients (17%), Raynaud phenomenon in 7 of 12 patients (58%), and fever in 5 of 12 patients (42%). Nearly all patients had radiographic features of usual interstitial pneumonia (UIP) on high-resolution chest computed tomography with bibasilar reticular infiltrates, traction bronchiectasis, and subpleural honeycomb change. One patient was deceased, but all living patients had active disease, mostly manifested as ILD. Three patients had undergone therapeutic lung transplant procedures during their follow-up period; all are alive, having survived a mean of 7 years after transplant.

Histopathologic and radiographic features are summarized in Table 2. Usual interstitial pneumonia was the most common pattern of lung involvement and was noted in 8 of 12 patients (67%) (Figure 1, A and B). Findings included typical, peripherally accentuated fibrosis of lung lobules with honeycomb changes and fibroblastic foci. Interstitial lymphoplasmacytic inflammation and lymphoid aggregates, with or without germinal centers, were seen in 8 of 12 patients (67%) (Figure 1, C). Two of 12 patients (17%) showed nonspecific interstitial pneumonia (NSIP) as the predominant finding (Figure 1, D). One of those patients had superimposed, presumably infectious, bronchopneumonia. Organizing pneumonia was the predominant pattern of lung disease in 2 of 12 patients (17%) (Figure 2, A), and a focal secondary finding in one patient with UIP and one patient with NSIP. Three of 12 cases (25%) showed a few scattered, nonnecrotizing granulomas (Figure 2, B): 1 with accompanying, presumably infectious, acute bronchitis, and another with features of aspiration. In the third case, granulomas were superimposed on UIP and were of uncertain etiology. A few scattered eosinophils were present in 10 of the 12 patients (83%), appearing prominently in one NSIP biopsy (Figure 2, C) and one patient with UIP and eosinophilic pleuritis. A pulmonary vasculopathy was apparent in 3 patients. In 2 cases, it was characterized by intimal fibrosis and medial hypertrophy of small- to medium-size arteries. In the third case, scattered small arterioles showed concentric muscular hypertrophy (Figure 2, D). Pertinent negative findings were the absence of vasculitis, necrotizing granulomas, amyloid deposits, and acute pleuritis in all cases.

Interstitial lung disease is a common feature in IIM. The reported incidence in DM and PM ranges from 20% to 65%, depending on the criteria used.¹¹  Nonspecific interstitial pneumonia is 4 times more common than UIP in patients with PM, whereas the difference is less than 2-fold in patients with DM, demonstrating the heterogeneity among patients with IIM in the degree and pattern of lung involvement.¹¹  Historically, the pattern of disease predicts prognosis: organizing pneumonia offers a better prognosis than UIP does, whereas diffuse alveolar damage portends a particularly poor prognosis.¹² 

Herein, we describe the histopathologic features of lung involvement in 12 cases of anti–PL-12 ARS, a distinct subset of IIM, although anti–PL-12 can be seen in clinical association with other CTD or UCTD. Our study provides 2 main findings: (1) there are no pathognomonic features that would allow for a definitive diagnosis of anti–PL-12 ARS based on the histopathologic features alone, and (2) UIP is the most common form of ILD found in this autoantibody cohort.

As noted by other authors, patients with autoimmune disease do not have histologically distinct lung biopsies of their ILD. However, according to Song et al,¹³  lymphoid aggregates and subjective, highly cellular biopsies are more often seen in this subgroup. Our cohort of patients with anti–PL-12 ARS mirrors their experience, with more than one-half of patients having a noticeable increase in lymphoid aggregates in their biopsies. Fibroblastic foci were not prominent in our cases, in keeping with results by Enomoto et al¹⁴  in their comparison of idiopathic UIP with CTD-related UIP.

Three of our 8 UIP specimens (38%) showed large areas resembling NSIP. Finding NSIP in patients with diffuse ILD usually raises the possibility of underlying, possibly occult, CTD because NSIP is a more commonly found pattern in CTD-related ILD. Several small series have identified NSIP as the most common pattern in patients with myositis-associated ILD.¹⁵  The prevalence of antisynthetase antibodies in these cohorts is either not known or not correlated with the histopathologic pattern. Heterogeneous ILD with an NSIP pattern has also been described in patients with idiopathic pulmonary fibrosis.¹⁶  At least a subset of NSIP cases appears to have the biologic potential to progress to a more advanced pattern of fibrosis.¹⁷ 

Nonnecrotizing granulomas were seen in 3 patients: 1 patient carried a diagnosis of Sjögren syndrome, the other 2 met criteria for UCTD, including 1 with elements of rheumatoid arthritis. Although granulomas have been associated with systemic conditions, such as rheumatoid arthritis, inflammatory bowel disease, and hypersensitivity reactions, they may also occur in the setting of infection, aspiration, tissue breakdown, or with certain pneumotoxic drugs. Their presence is not specific enough to strongly suggest lung involvement by ARS.

Histologic evidence of a vasculopathy was observed in 3 of 12 patients (25%) in this cohort. In 2 patients, this finding correlated with increased mean pulmonary artery pressures of 48 and 27 mm Hg, respectively, as determined by right heart catheterization. The third patient had no evidence of pulmonary hypertension by echocardiography. Pulmonary hypertension is a known complication of ARS. Kalluri et al⁵  observed pulmonary hypertension, inferred by means of echocardiography, in 9 of 31 patients (29%) possessing anti–PL-12 antibodies.

Table 3 summarizes the pulmonary histopathology of 46 anti–PL-12 cases in the English literature to date. Although the nomenclature is not uniform in these reports, about two-thirds of patients seem to have some form of cellular and/or fibrosing ILD. Follow-up was not available for many cases. The existing literature and our cohort confirm that diffuse alveolar damage is not only uncommon in IIM, in general, but also in anti–PL-12 ARS specifically.^18,19  The only patient with diffuse alveolar damage reported in the literature also showed capillaritis and was found to have improved on glucocorticoids 8 months after the episode. One patient who died rapidly of respiratory failure may have had an acute lung injury resembling diffuse alveolar damage.³ 

Distinguishing ILD arising in the setting of ARS from idiopathic variants of ILD is important because recent data show that survival by patients with myositis-associated UIP is better than a matched cohort of patients with idiopathic pulmonary fibrosis.²⁰  Even within the group of ARS, patients with anti–Jo-1 autoantibodies seem to fare better than do those with other anti-tRNA synthetase autoantibodies, such as anti–PL-12, predominantly because of worse pulmonary outcomes.⁷  Thus, considering an ARS in the setting of isolated ILD is valuable because lung disease is the presenting feature in a significant number of such patients.⁵  In rare cases, antisynthetase autoantibodies coexist with other autoantibodies, especially Sjögren syndrome–associated antibody anti-SSA. The presence of anti-SSA in patients with ARS may be associated with worse pulmonary outcomes.^4,5,21  Moreover, ARS could be considered and serologically assayed, even in patients with an established autoimmune disease other than myositis.

This study has limitations. We are a tertiary referral center for ILD and rheumatic diseases, as well as a lung transplant center. Our patient cohort may be skewed toward late-stage and progressive lung diseases, such as UIP. However, most of the patients in this study lived close to our center and were followed by rheumatologists within our medical center before referral for lung transplantation. Although UIP representing idiopathic pulmonary fibrosis in patients older than 50 years cannot be excluded with absolute certainty, we believe that the patients in this series most likely suffered from lung involvement by ARS rather than from idiopathic disease because of the high prevalence of ILD in patients with ARS. Furthermore, our patients often experienced the onset of respiratory symptoms long before the diagnosis of ARS and lung biopsy.⁷  Two of our patients showed histopathologic features of NSIP in biopsy material. We cannot exclude that these patients were inadequately sampled, especially because the radiologic diagnosis of UIP has high predictive value.²²  However, both biopsies were taken from the lower lobe, making missing UIP less likely if it were present. Finally, a limitation of this study, as seen with any study of this kind, is the lack of an age-matched comparison group showing idiopathic disease.

In summary, there are no pathognomonic histopathologic features to distinguish anti–PL-12 ARS-related lung disease from idiopathic variants of diffuse ILD. Clinical features of mechanic's hands, Raynaud phenomenon, arthritis, and fever should prompt pathologists to suggest involvement by ARS. Because the prognosis of IIM-associated UIP appears to be better than that of idiopathic UIP treated with systemic immunosuppressive therapy, considering the possibility of lung involvement by ARS is important.