Concomitant Interstitial Lung Disease with Psoriasis

Background We encounter interstitial lung disease (ILD) patients with psoriasis. The aim of this case series was to examine clinical and radiographic characteristics of patients with concomitant psoriasis and ILD. Methods This is a retrospective review of our institutional experience of ILD concomitant with psoriasis, from the database in the Advanced Lung/Interstitial Lung Disease Program at the Mount Sinai Hospital. Out of 447 ILD patients, we identified 21 (4.7%) with antecedent or concomitant diagnosis of psoriasis. Clinical, radiographic, pathological, and outcome data were abstracted from our medical records. Results Median age was 66 years (range, 46–86) and 14 (66.7%) were male. Thirteen (61.9%) had not previously or concomitantly been exposed to immunosuppressive therapy directed against psoriasis. Two (9.5%) ultimately died. Clinical diagnosis of ILD included idiopathic pulmonary fibrosis, 11 (52.4%); nonspecific interstitial pneumonia (NSIP), 2 (9.5%); cryptogenic organizing pneumonia, 2 (9.5%); chronic hypersensitivity pneumonitis, 2 (9.5%); and the others, while radiographic diagnosis included usual interstitial pneumonia pattern, 9 (42.9%); NSIP pattern, 6 (28.6%); organizing pneumonia pattern, 4 (19.0%); hypersensitivity pneumonitis pattern, 2 (9.5%); and the others. Conclusions We report 21 ILD cases with antecedent or concomitant diagnosis of psoriasis. Further prospective studies are required to determine the association between ILD and psoriasis.


Introduction
Psoriasis commonly presents with sharply defined erythematous plaques with overlying silvery scales. e scalp, elbow extensors, knees, and back are common locations for plaque psoriasis [1].
is hyperproliferative state is characterised by increased numbers of epidermal stem cells and cells undergoing DNA synthesis, shortened keratinocyte cell cycle time, and decreased epidermal turnover time [1]. Psoriatic arthritis is an inflammatory arthritis/synovitis that occurs in 25% of patients with psoriasis; it is a distinct clinical entity characterised by involvement of the distal interphalangeal joints of the hands and feet as well as the absence of rheumatoid factor [2]. Previous studies demonstrate that psoriasis is associated with diabetes mellitus [3], arterial hypertension [4], obesity [5], dyslipidaemia [6], cardiovascular disease [7], and nonalcoholic fatty liver disease [8].
Interstitial lung disease (ILD) is defined as any lung disease occurring in the parenchymal interstitium (i.e., the alveolar wall or alveolar septa) or loose-binding connective tissue (i.e., the peribronchovascular sheath, interlobular septa, or pleura) [9]. Type 17 helper T (T(H)17) cells were reported to be one of the common pathways which contribute to alveolitis and enhance cytokine production in pulmonary fibroblasts. T(H)17 cells are also the hallmark of pathogenesis in psoriasis; therefore, the coexistence of ILD and psoriasis is worth exploring on the basis of possibly shared underlying immunologic pathways. A previous small study (N � 50) in the Tucson VA Hospital showed that pulmonary fibrosis was not more common in a population of patients with psoriasis than would be expected in a control VA population [10]. ILD is occasionally reported in patients with psoriasis, most cases of which have been reported previously as drug-induced pneumonitis secondary to concomitant use of immunosuppressants. However, in the ILD program at Mount Sinai Hospital, we frequently encounter ILD patients with a concomitant diagnosis of psoriasis, who had been free from immunosuppressive therapy directed against psoriasis. In this pilot case series, we reported our experience with ILD in patients with psoriasis to assess their clinical features and outcomes.

Patients.
is case series was a retrospective review of our institutional experience with ILD concomitant with psoriasis at the Advanced Lung/Interstitial Lung Disease Program at Mount Sinai Hospital, New York. e medical records of 447 consecutive patients who visited the ILD program were reviewed retrospectively. e inclusion criteria were as follows: (1) patients who visited the Advanced Lung/Interstitial Lung Disease Program from October 2009 through September 2015, (2) age ≥ 18 years old, (3) a diagnosis of ILD, and (4) diagnosed with either psoriasis or psoriatic arthritis by dermatologists. Patients who had ILD associated with sarcoidosis were excluded. Of the 447 patients who visited the ILD program, 21 (4.7%) had a diagnosis of both ILD and psoriasis. All findings on computed tomography (CT) were reviewed by a board-certified radiologist (MS), and pathological specimens were reviewed by a board-certified pathologist (MBB). All clinical and outcome data of the 21 patients were abstracted from our medical records. e following data were extracted: age, sex, symptoms, time from diagnosis of ILD (months) prior to initial enrolment, time from diagnosis of psoriasis (years), comorbid illnesses (i.e., other autoimmune diseases, gastroesophageal reflux, and pulmonary hypertension (PH)), body mass index, smoking history, New York Heart Association (NYHA) functional class, medications, home oxygen therapy, pulmonary function test, 6-minute walking test, clinical diagnosis of ILD, CT findings, pathological results, and follow-up information. We defined both radiographic "definitive UIP pattern" and "probable UIP pattern" as "UIP pattern" in the present study. Echocardiographic evidence of possible PH was defined as estimated pulmonary artery systolic pressure >40 mmHg.
Our ILD program involves weekly multidisciplinary discussions with board-certified pulmonary physicians (AM, DS, and MLP), radiologist (MS), and pathologist (MBB), who make clinical diagnoses based on the clinical, radiographic, and pathological findings of each ILD case.

Statistical Methods.
Descriptive analyses were performed. Continuous variables are presented as median and range, and categorical variables are presented as frequency and percentage unless stated otherwise.

Patient Characteristics and Clinical and Pathological
Presentation. e characteristics of the 21 patients with ILD and an antecedent or concomitant diagnosis of either psoriasis or psoriatic arthritis are shown in Table 1; their demographic and clinical features are shown in Table 2. Median patient age was 66 years (range, 46-86 years). ere were 14 (66.7%) men and 7 women. ere were 15 (71.4%) current or former smokers. e most common symptom on presentation was "cough," followed by "dyspnea on exertion." e median time from ILD diagnosis prior to initial enrolment was 24 months (range, 0-132 months), while the median time from psoriasis diagnosis prior to the enrolment was 2 years (range, 1-22 years). Most patients maintained relatively good performance status (NYHA I/II: 19 (90.5%)).

Discussion
To our knowledge, the present case series is the first to fully investigate the clinical and radiographic features of ILD with ANA: antinuclear antibody, ANCA: antineutrophil cytoplasmic antibody, CHP: chronic hypersensitivity pneumonitis, COP: cryptogenic organizing pneumonia, CTD-ILD: connective tissue disease-associated interstitial lung disease, EP: eosinophilic pneumonia, F: female, GGO: ground glass opacities, HP: hypersensitivity pneumonitis, IPF: idiopathic pulmonary fibrosis, M: male, MP: mercaptpurine, MTX: methotrexate, NSIP: nonspecific interstitial pneumonia, OP: organizing pneumonia, PE: pulmonary embolus, PM/DM: polymyositis/dermatomyositis, RBILD/DIP: respiratory bronchiolitis-associated interstitial lung disease/desquamative interstitial pneumonia, SSA: SS-A antibody, and UIP: usual interstitial pneumonia. * A staging system, originally proposed for ILD with systemic sclerosis, by Goh et al. [11]. * * Patient had a mild elevation of serum SS-A (1.20 AI) but lacked sicca symptom and other symptoms suggestive of underlying autoimmune disease (i.e., arthralgia, digital fissuring/ulceration, and Raynaud's phenomenon). * * * Patients had positivity of serum ANCA but lacked clinical symptom suggestive of underlying systemic vasculitis (i.e., skin rash, alveolar haemorrhage, and glomerulonephritis). # We defined both radiological "definitive UIP pattern" and "probable UIP pattern" as "UIP pattern" in the present study. psoriasis. Of 447 patients who visited the ILD program in a referral practice, 21 ILD patients (4.7%) had concomitant psoriasis or psoriatic arthritis, and most of them (63.6%) had not been previously or concomitantly exposed to immunosuppressants. In previous studies (Table 3), the following immunosuppressants were associated with pneumonitis in patients with psoriasis: methotrexate [32], infliximab [26], fumaric acid ester [25], leflunomide [24], gold [18], and sulfasalazine [17]. In contrast, only 3 cases of ILD without concomitant or prior use of immunosuppressants have been reported. Messina et al. [21] report a case of psoriasis with a pathologically OP pattern, even though concomitant cytomegalovirus infection might have induced pneumonitis; therefore, the causality is unclear in this case. Meanwhile, Hiki et al. report a case of severe IgA nephropathy associated with psoriatic arthritis and idiopathic interstitial pneumonia [13]. ird, Gupta and Espiritu showed a case of worsening exertional dyspnea, found to have NSIP pattern in CT chest, and subsequently developing punch biopsy-proven psoriasis [29]. Our institutional experience with 21 ILD/psoriasis cases, most of which had been without concomitant or prior use of immunosuppressants, potentially suggests a possible direct association between these 2 different disease entities. e previous small study (N � 50) in the Tucson VA Hospital failed to reveal their correlation, and perhaps, this might have been attributed to the underpowered nature of the study (i.e., insufficient sample size) [10]. erefore, prospective studies with sufficient sample size, involving careful assessment with pulmonary function testing and high-resolution CT in a psoriasis cohort, are imperative for accurately determining the prevalence of ILD and assessing their possible association. e relation between psoriasis and chronic obstructive pulmonary disease was recently demonstrated [33]. Due to its possible autoimmune-related pathogenesis, psoriasis-related immune dysfunction could lead to an abnormal immunologic response in the lung parenchyma. We hypothesise that immune dysfunction in patients with psoriasis similarly causes inflammation and fibrotic process in the lung interstitium. Activated T(H)17 cells were recently found to produce several mediators such as interleukin 17A, 17F, and 22, which induce keratinocyte proliferation and other hallmark features of psoriasis [34]. Interestingly, interleukin 17, which is produced by T(H)17 cells, contributes to alveolitis and enhances cytokine production of pulmonary fibroblasts [35]. However, the future immunology research is crucial to confirm our speculation that the pathogenic role of interleukin 17 is equivalent in the skin and lungs. Another study reports that dermal cell proliferation in psoriasis is secondary to activated TGF-α, which may also be associated with activated TGF-β, leading to pulmonary fibrosis [19]. Moreover, abnormal histologic features of lung tissue were observed in 4 (44.4%) out of 9 patients with rheumatoid factor-negative arthritis (2 with fibrosis, 1 with follicular lymphoid hyperplasia, and 1 with DIP) who did not have respiratory-related symptoms [36]. ese findings collectively suggest that the coexistence of ILD and psoriasis is mechanistically plausible. e pulmonary disease in psoriasis could potentially be similar to that in ankylosing spondylitis (AS) because of their clinical similarities (negative rheumatoid factor) between psoriatic spondylitis and AS. However, the incidence of pulmonary involvement in AS is also uncommon; a previous review at the Mayo Clinic suggests that the frequency is <2% [37]. e patients in the present study showed relatively mild impairment in FVC% predicted (median, 69; range,  in contrast to a severely decreased DLco% predicted (median, 47; range, 17-95). Given the high prevalence of concomitant PH or possible PH (42.9%) in our patients, this discrepancy between FVC and DLco might result from disproportionally elevated pulmonary artery pressure. Expectedly, mild PH is reported to be significantly more frequent in psoriasis patients, implying that increased antigen presentation, increased cutaneous T-lymphocyte activity, interleukins, and tumour necrosis factor-α in psoriasis cause endothelial dysfunction, which is an important pathophysiological characteristic of PH [38]. erapeutic modalities for PH triggered by psoriasis should be investigated further.
In the present study, 11 (52.4%) patients had a clinical diagnosis of IPF; regarding radiographic findings, the UIP pattern was more common (42.9%) than the NSIP/OP pattern. is might suggest that immune dysfunction in the lungs triggered by psoriasis tends to cause fibrotic change rather than the conventional inflammatory process. e findings in our patients resemble the patterns observed in rheumatoid arthritis, in which a radiographic UIP pattern is the most common. erefore, these 2 different diseases might have a common respiratory pathophysiology in setting of a shared clinical manifestation such as joint synovitis. Furthermore, compared to conventional IPF patients, the decline of pulmonary function was relatively mild, as indicated by a modest change in the rate of predicted FVC% (range, − 5.7% to +18%/year). is also implicates a distinct feature of IPF associated with psoriasis as compared to conventional IPF. In terms of ILD treatment, Gupta and Espiritu [29] reported of a patient with NSIP associated with psoriasis responding to azathioprine, while Miyachi et al. [30] demonstrated a case which highlights the improvement of the interstitial lung pattern during psoriasis treatment with secukinumab. ese two reports anecdotally support the efficacy of immunomodulatory therapy for ILD in a patient with underlying psoriasis. e efficacy of antifibrotic agents (i.e., pirfenidone and nintedanib) for psoriasis-associated IPF would be worth exploring in the future.
e major limitation of this study is that the direct association between psoriasis and ILD cannot be concluded from the findings of this case series, although the prevalence of psoriasis in our ILD cohort was not negligible. Due to the retrospective nature, the study information was loaded with several confounding factors (varied ILD patterns, smoking status, use of immunosuppressants, and other autoimmune features) that make it difficult to draw a valid conclusion. For example, the high prevalence of smoking status (71.4%) in our cohort might have been accounting for increased incidence of ILD as smoking was known to be one of the risk factors for IPF [39]. Prospective studies comparing prevalence of ILD in the psoriasis cohort and that in the general population are required to delineate Canadian Respiratory Journal

Conclusion
We reported 21 ILD cases with antecedent or concomitant diagnosis of psoriasis. Among our ILD cohort, 4.7% of those had concomitant diagnosis of psoriasis or psoriatic arthritis. Considering the 3.2% prevalence of psoriasis among adults [40], this might suggest real association between two different disease entities, which cannot be concluded by this case series due to the major limitation discussed above. Further prospective and large clinical studies are warranted to investigate the association between these two different disease entities.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Disclosure
An abstract reporting preliminary data of this work was presented at the 2017 American oracic Society Annual Meeting, May 24, 2017, Washington, DC.

Conflicts of Interest
Mary Salvatore is a speaker for Genentech and Boehringer Ingelheim. e rest of the authors do not have any conflicts of interest to declare.