Pulmonary Involvement in Microscopic Polyangiitis: Computed Tomography Findings in 55 Patients With Analysis of Risk Factors for Recurrence

Background and objective Pulmonary involvement is seen in up to 30% of microscopic polyangiitis (MPA) patients. Pulmonary radiological findings for MPA have been scarcely reported to date. This study was conducted to evaluate computed tomography (CT) and clinical findings at the time of MPA diagnosis as predictors for systemic or lung recurrence. Methods We retrospectively reviewed the medical records and radiological data of 55 MPA patients with pulmonary involvement who were admitted to our hospital between April 2008 and December 2016. Results Aside from pulmonary lesions, lesions were found in the kidneys (52.7%), skin (7.3 %), and peripheral nerves (3.6%). Biopsies were performed for 29.1% of the patients, with an overall diagnostic accuracy of 78.9%. Parenchymal opacities (74.5%, mainly ground-glass opacities and reticular shadowing) were more commonly seen than airway abnormalities were (40.0%, mainly bronchiectasis). Systemic recurrence in the first year after diagnosis was found in 10.9% of the patients, and it mainly involved the kidneys or lungs. A serum WBC count ≥ 10,900/μL was a risk factor for predicting systemic recurrence within the first year after diagnosis according to the Cox regression analysis (HR 11.1, 95%CI: 1.3-95.9, p=0.028). Lung recurrence within five years after the diagnosis was observed in 9.1% of the patients. The incidences of reticular shadowing and honeycombing in thoracic CT at diagnosis were significantly higher in recurrence-positive patients than in recurrence-negative patients, but these differences could not be used to predict lung recurrence. Conclusions Ground glass opacities, reticular shadowing, and bronchiectasis are prominent thoracic CT findings for MPA. There are no radiological patterns capable of predicting recurrence. However, a serum WBC count ≥ 10,900/μL at diagnosis might be a predictive factor for systemic recurrence within the year.


Introduction
Microscopic polyangiitis (MPA) is a systemic necrotizing vasculitis that affects small-caliber blood vessels, sometimes including those of the lungs. Previous studies have reported the frequency of lung involvement at 22%-29% [1] To the best of our knowledge, the clinical and radiological features of MPA have not yet been clearly described, and the risk factors for systemic or lung recurrences of MPA are unknown. This study used clinical observations and radiological findings derived from thoracic computed tomography (CT) scans to investigate whether data available at the time of diagnosis of MPA can be used to predict the likelihood of recurrence.

Patients and study design
We retrospectively reviewed the medical records of MPA patients admitted to our hospital, a regional referral center, between April 2008 and December 2016.
We enrolled 55 patients with lung involvement who satisfied the criteria for MPA given in Japanese Labour and Welfare, Japan 1998 but excluding patients who had no lung lesions detectable in thoracic CT scans, or for whom CT scans were not available (n=8), or who in addition to MPA had other lung conditions (n=5), such as lung infections (nontuberculous mycobacterium, lung abscesses) or collagen-related vascular diseases (Sjogren syndrome plus Scleroderma, Rheumatoid arthritis) ( Figure 1). The clinical and thoracic CT findings were evaluated at the time of diagnosis. Three pulmonologists, each with more than 10 years of experience, independently reviewed the high-resolution CT findings; decisions were interpreted by consensus. The thoracic CT findings were divided into two general categories: 1) airway abnormalities (bronchial wall thickening, bronchiectasis, centrilobular nodules); 2) parenchymal opacities (ground-glass opacities [GGO], consolidation, reticular shadowing, nodules, cavities, hemorrhage, honeycombing). Figure 2 illustrates representative thoracic CT findings.

Comparison of patients with and without systemic recurrence in the year following diagnosis
We compared clinical and radiological findings for the two groups defined according to whether there was systemic recurrence in the first year after the initial MPA diagnosis. Systemic recurrence was defined as the occurrence of one or more of the following events: 1) a newly observed increase in myeloperoxidaseantineutrophil cytoplasmic antibody (MPO-ANCA) titer, 2) progression of previously observed organ dysfunction, and 3) emergence of involvement of other organs.

Comparison of patients with and without lung recurrence in the fiveyear period following diagnosis
We also compared clinical and radiological findings for groups defined according to whether there was recurrence in the lungs in the five-year period after the diagnosis of MPA. We evaluated the findings available at the time of MPA diagnosis as predictive factors for recurrence in the lungs.

Statistical analysis
The data were analyzed statistically using a Pearson's chi-squared test or Mann-Whitney U test. Statistical Package for the Social Sciences (SPSS) Statistics (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY: IBM Corp.) was used for all analyses. The overall survival and recurrence rates were evaluated using the Kaplan-Meier method. A log-rank test was used to statistically compare the curves and p values. The cut-off point for the serum markers was determined as the minimum value based on the formula of [(1-sensitivity)2+(1-specificity)2]. A p-value < 0.05 in paired two-sided tests was considered significant. This study was approved by the ethics committee of Kyorin University (approval number 889 H28-149, dated Feb 10, 2017).

Results
We identified a total number of 68 MPA patients over the study period. Eight were excluded owing to the absence of thoracic CT scans, and another five were excluded owing to complications of lung involvement by co-existing collagen vascular diseases such as rheumatoid arthritis (n=1), systemic scleroderma combined with Sjögren syndrome (n=1), lung abscesses (n=1), or non-tuberculous mycobacterium infections (n=2). The remaining 55 patients were enrolled in this study ( Figure 1).
Approximately half of the patients (n=27, 49.1%) were treated using steroid pulse therapy.

Radiological findings at the time of diagnosis for patients with and without lung recurrence within five years
The proportions of patients showing airway abnormalities or parenchymal opacification at the time of diagnosis were similar between the groups who did or did not show lung recurrence in the following fiveyear period (

Comparison of the overall survival with or without recurrence in any organ
Using the Kaplan-Meier method, survival rates between the two groups with or without recurrence in any organ did not differ in the following period: one year (p=0.169), two years (p=0.843), and five years (p=0.843) after diagnosis. Similarly, the one (p=0.508), 2 (p=0.374), and five years (p=0.346) survival rates were comparable between the lung recurrence positive and negative groups.

Discussion
Owing to the rarity of MPA, few reports of radiological findings have been published [2,3]. Therefore, the identification of radiological and laboratory findings available at the time of diagnosis that is predictive of recurrence has been difficult. The present study is unique as radiological findings could be analyzed and the frequency of recurrence up to five years after the diagnosis of MPA was examined, permitting analysis of the factors that are relevant to recurrence. The data supports the following conclusions: Based on radiological observations, the most frequently observed types of parenchymal opacities were ground-glass opacities and reticular shadows, while the most frequently observed airway abnormality was bronchiectasis. The CT findings for MPA were diverse in this study, but not substantially different from those previously reported [2,3].
The incidence of systemic recurrence in the first year after diagnosis was 10.9% (n=6), and the kidneys and lungs were the most frequently involved organs. Previous reports have described that the presence of antineutrophil cytoplasmic antibody (ANCA) in serum was a better predictor of lung fibrosis than the presence of proteinase-3 (PR3)-ANCA was, but it did not affect the recurrence of ANCA-associated vasculitis. This was also seen in our study. Furthermore, no predictor has been shown to reliably guide therapeutic decision-making [4,5]. Hatemi et al reported that an increase of PR3-ANCA titer may help to predict relapses in ANCA-associated vasculitis patients and PR3-ANCA may also be associated with better response to rituximab [6]. Similarly, Schirmer et al. [7] analyzed PR3-ANCA positive granulomatosis with polyangiitis (GPA), myeloperoxidase (MPO)-positive GPA, and MPO-ANCA positive MPA and found that increased relapse rate was observed among patients with PR3-ANCA irrespective diagnosis. While, this study demonstrated that patients with recurrence had significantly higher serum WBC counts than patients without recurrence did, and high WBC counts (≥ 10,900/μL) at the time of diagnosis might be a potential marker for predicting systemic recurrence within the first year.
Lung recurrence occurred in 9.1% of MPA patients within five years of diagnosis. Although the incidence of reticular shadowing or honeycombing in thoracic CT scans at diagnosis was significantly higher in recurrence-positive patients than in the negative patients, these CT findings could not predict lung recurrence according to the Cox regression analysis. A previous report suggested that pulmonary interstitial fibrosis, including honeycombing, can be an early manifestation of MPA, preceding the onset of vasculitis by several years [3]. However, the present study clearly demonstrated that fibrosis by itself cannot predict lung recurrence within five years after diagnosis. The risk factors for lung recurrence are therefore numerous, implying that it will be difficult to predict lung recurrence based on only laboratory and radiological findings at diagnosis.
Furthermore, this study demonstrated that neither lung nor systemic recurrence involving any organ seemed to affect the overall survival; however, more index cases will be needed to confirm these results.
This study illustrates MPA as a systemic disease that most frequently involves the lungs and kidneys. Only 29.1% of patients were examined pathologically, but the diagnostic accuracy in these cases was 78.9%. This result should not be interpreted as an assurance of diagnosis based on tissue biopsy of any organ, because a lung biopsy showed only necrotizing vasculitis affecting small vessels including arterioles, venules, or capillaries, even in autopsied specimens, as previously reported [8]. This suggests that pulmonologists should avoid using lung biopsies to diagnose MPA.
This study has some limitations. First, it was a retrospective study involving a small number of MPA patients. Second, all the enrolled patients showed pulmonary involvement on thoracic CT, which constitutes a selection bias. However, this was a necessary price to pay in order to make use of information derived from thoracic radiological findings to predict systemic and lung recurrence.
However, this study demonstrated that thoracic involvement in MPA was primarily characterized by parenchymal opacities (n=40, 74.5%), such as ground-glass opacities and reticular shadowing, and to a lesser degree by airway abnormalities (n=22, 40.0%), mainly presenting as bronchiectasis. None of the laboratory or radiological findings available at the time of diagnosis could predict lung recurrence during the following five years. However, serum WBC count (≥ 10,900/μL) at the time of diagnosis could be a potential marker for predicting systemic recurrence within the first year.

Conclusions
Thoracic involvement in MPA was primarily characterized by parenchymal opacities (ground-glass opacities and reticular shadowing) followed by airway abnormalities mainly presenting as bronchiectasis. Although none of the laboratory or radiological findings available at the time of diagnosis could predict lung recurrence during the following five years, serum WBC count (≥ 10,900/μL) could be a potential marker for predicting systemic recurrence within the first year.