Quantitative High-resolution Computed tomography Analysis of Antisynthetase Syndrome with Interstitial Pneumonia

Background Non-specic interstitial pneumonia (NSIP) combined with organizing pneumonia (OP) pattern has been conrmed in pathology or radiology. This study is to analyze the correlation between high-resolution computed tomography (HRCT) quantitative indexes (QI) and pulmonary function test parameters (PFTs), and compare differences of HRCT patterns in Antisynthetase Syndrome with interstitial pneumonia (ASS-IP). Methods Data of ASS-IP patients admitted to respiratory department of Ping Jin hospital from January 2014 to December 2019 were retrospectively reviewed. Results 21 ASS-IP patients were enrolled in this study and 3 patients were NSIP, 9 patients were OP and 8 patients were NSIP overlap OP (NSIP/OP) pattern. (1) Volume, Weight percentages of the extracted whole lung volume with attenuation values (V%, W %) and mean lung attenuation (MLA) of non-aerated area is 5.69%, 16.43% and -11.76 Hounseld unit (Hu). Total MLA of attenuation values(MLA total ) is -678.97 Hu. (2)FVC and MLA of poorly aerated (cid:0) r=0.58,P=0.048 (cid:0) , FEV 1 and MLA (cid:0) r=0.79,P=0.001 (cid:0) of poorly aerated, FEV 1 /FVC (cid:0) r=0.23,P=0.01 (cid:0) and MLA of normally lung aerated, have linear correlation .(3) DLCO (P=0.049), volume of poorly aerated (V brosis %,P =0.03), and weight of hyper inated (W hyper %,P=0.02) percentages, and MLA total (P=0.01) have signicant statistic differences between patients with NSIP/OP or OP patterns, but the therapy improvement time(P=0.41) had no difference. partly classication ASS-IP. NSIP/OP brosis %, W brosis %,W normal % and W hyper %,Weight percentages of the extracted whole lung volume with attenuation values of non-aerated area, poorly aerated and brosis ,normally lung aerated(normal) and hyper inated.MLA, mean lung attenuation. MLA nonaerated , MLA brosis , MLA normal, MLA hyper and MLA tota , MLA of the extracted whole lung volume with attenuation values of nonaerated area, poorly aerated and brosis ,normally lung aerated,hyper inated area and outlined total lung area. ET, Effective time of treatment.

Conclusions QI of HRCT has a good correlation with PFTs and partly reference value to pathological classi cation of ASS-IP. NSIP/OP may be a common pattern in ASS-IP, brosis severity of which is more severe than OP. Background Antisynthetase syndrome (ASS) is a subtype of Dermatomyositis and Polymyositis (DM/PM), which has characteristic changes, that together with interstitial pneumonia (IP), in ammatory arthritis, fever, Raynaud phenomenon and mechanic's hands [1].Serum anti-Jo-1 is the most frequent positive antibody in a patient. However, other speci c anti-Antisynthetase such as anti-alanyl (PL-12), anti-threonyl (PL-7), and anti-glycyl (EJ) tRNA-synthetase antibodies have been found routinely [2].In addition, some previous autoimmune antibodies have also been reported to associate with the development of ASS with interstitial pneumonia (ASS-IP), such as anti RO-52 [3].
Non-speci c interstitial pneumonia (NSIP) or organizing pneumonia (OP) is pathology or radiology patterns, pathogenic factors of which included idiopathic or secondary forms. High-resolution computed tomography (HRCT) pattern of connective tissue diseases (CTDs) associated with IP can be classi ed by the criterion of idiopathic IP guidelines [4]. Pathological studies of lung brosis detected some biopsies shown NSIP overlap OP (NSIP/OP) along bronchial blood vessel bundle with bilateral the lower lobes subpleural region predominance [5][6][7] (4,5,6).It reported that NSIP or NSIP/OP pattern is accounted for 74-90% [8,9] in DM/PM. HRCT can provide classi cation of IP patterns and assessment of IP severity and therapeutic guide.
Computer-aided quantitative analysis (QA) is substituted visual appraisal due to its accuracy and objectivity, of which the threshold mask method is more popular because of its availability [10]. In this article, we analyzed clinical character of ASS-IP patients and were looking for HRCT quantitative associations with clinical pulmonary functions test parameters (PFTs).

Patient
We performed a retrospective review data of antisynthetase antibodie positive patients admitted to respiratory department of Ping Jin hospital from January 2014 to December 2019. Patient diagnosed as ASS-IP met IP guideline [11] and the criterion proposed by Connors colleagues [12,13].Collected and recorded the basic information of patients recruited, laboratory test results, treatment strategies and follow-up data respectively.
HRCT scan and QA HRCT imaging was performed with a 64 section mufti-detector CT scanner (GE Healthcare, Milwaukee, WI). The QA of HRCT was obtained using a commercially available workstation by two chest radiologists (Zhao da Wei and Yin liang) with 10 years' experience. The standard based by characters of signs and distribution of HRCT and criteria recommended in previous literature [14,15].In each slice, the operator manually outlined the lung borders through selecting the total lung tissue and excluding the heart, the central airways, and the major blood vessels. Total lung volume (TLV) was described as lung parenchymal volume of outlined lung area .TLV and mean lung attenuation (MLA) can be calculated by Workstation. Attenuation area ranges were de ned differently aerated lung compartments: non aerated,-100 to +100 Houns eld unit (Hu), which represent the areas of pulmonary consolidation [16]; poorly aerated and brosis ( brosis),-101 to -500 Hu, which represent the areas of ground-glass opacity and reticulation; normally lung aerated(normal),-501 to-900 Hu; and hyper in ated (hyper),-901 to-1,000 Hu [17,18].Volumes% and weight% of Attenuation area were calculated as percentages of the extracted whole lung volume with attenuation values ranging from 100 Hu to −1000 Hu [17,18].A μ value is assigned to each voxel (μ voxel), roughly proportional to its density, and is expressed as a CT number standardized to that of water. CT value = 1000 × μ voxel -μ water/μ water [16].Proportional weight=μ voxel × proportional Volumes.

PFTs
PFTs performed by a spirometer (JAEGER Master Screen) are used for diagnosis and monitoring of ASS-IP patients, including FVC, FEV1, FEV1/ FVC, carbon monoxide diffusion (DLCO).

Treatment strategies and the evaluation of effect
Glucocorticoids and immunosuppressant strategies of each patient were studied and therapeutic effect was evaluated by changes of FVC or DLCO [19].Moreover, if one patient had no HRCT in the follow up, we de ned patient as improvement case by 10 percent of V normal % over the baseline value and recorded the length of the treatment.

Statistical analysis
Patients' data were analyzed by Spss 19.0 statistical software. Their data were described as mean, interquartile range or as percentage of the relative frequency according to the different. Categorical data were compared with using the χ2 test or Fisher's exact probability test for independence. Correlation analysis between PETs and different differently aerated lung compartments quantitative parameters. P < 0.05 was considered as statistically signi cant.

Results
This study included a total of 21 ASS-IP patients and baseline clinical characteristics show Table 1.The median age at diagnosis was 51 years, 13 patients were female and 8 were smokers. Serum antibody test showed 4 patients were positive anti-PL-7, 7 patients were positive anti-Jo-1, 6 patients were anti-PL-12 and 3 patients were anti-EJ. There are 15 patients with cough, 11 patients with dyspnea, and 21 patients with crackles in the lower pulmonary lobes. Beside serum positive Anti-synthase antibody, autoimmune serology test was positive for ANA, rheumatoid factor IgM and antibodies to Ro-52.Arterial blood gas (ABS) indicated hypoxemia in most cases. Respiratory symptoms or CTDs related symptoms and physical ndings, PFTs and ABS were not different between two groups.

Treatment strategies and follow-up
All patients were given to treatment strategies with glucocorticoids combined with immunosuppressants.
During follow-up, 3 patients lost connections (1 patient died of respiratory failure after 20 months of treatment,2 patients were out of touch).The improvement time is taken from 1 to 6 months and it found that there is no different between NSIP/OP and OP patterns groups.
Discussion IP is one of the most common CTDs with high morbidity and mortality [20], while the ASS-IP is a special type of IP. The diagnosis of ASS-IP requires not only evidence of ILD, such as pathology and HRCT images, but also evidence of positive serum antibodies, just mentioned in methods. Hence that ASS-IP is also a rare type in clinic. In this study, we had retrospectively collected 23 cases of ASS-IP which were diagnosed and treated in our hospital in the past 5 years. All cases meet the diagnosis and exclusion criteria. Among the 23 patients, the mean age is 51 years old and female accounts for the most. Dyspnea and cough are common symptoms. Fine crackle can be heard in all patients. Mechanic's hand is major sign of rheumatic diseases. Previous research suggested that pulmonary arterial hypertension has a low prevalence in ASS-IP compared with other CTDs, but it independently affects prognosis and survival [21]. In our study, one out of the 23 patients had pulmonary arterial hypertension and died after 20 months follow-up in this study. Tacrolimus is main immunosuppressants used as the standardized treatment and the patients can improve after the standardized treatment for mean 3.44 months.
Interpretations from radiologists on HRCT were mainly based on visual assessment. Some investigators used semiquantitative way to evaluate severity and monitor progression of disease, and score the IP by the international Goh score with evaluation at 5 lung levels [22,23]. However, its result cannot represent the whole lung actuality because of limited levels analysis. QA of HRCT has been widely carried out in both researches and clinical practice in recent years. Through analyzing HRCT image, a serial of parameters could be acquired, such as average density, texture analysis and threshold segmentation in high attenuation areas Threshold segmentation software is popularly used by the operator at home and abroad because of easy availability. It can quantify interstitial lung changes by setting different threshold value range. Under functioning lung volume (-700 Hu > pixel > -950 Hu) and IP volume (>-700 Hu),quantitative indexes showed good correlation not only with the extent of IP estimated by visual inspection but also with PFT results [24].Another study found that when the high attenuation area (threshold value >-500 Hu) was de ned as follows IP, and percent of which were signi cantly correlated with DLCO [25].Considering the performance of the consolidation pattern of ASS-IP patient, the high attenuation area between − 100 to100 Hu is also analyzed and de ned as non-aerated areas in our research. Among pulmonary function parameters and aerated lung compartments quantitative values, FVC% and MLA brosis (r = 0.58,P = 0.048),FEV1/FVC and MLA normal (r = 0.23,P = 0.01),and FEV1% and MLA brosis (r = 0.79,P = 0.001)have good linear correlation .CTD associated IP often present either an NSIP or OP pattern. NSIP/OP pattern has been con rmed by some clinicopathological studies [26],which was reported to be associated with CTDs [5].Therefore, we focused on comparing quantitative indexes between NSIP/OP and OP pattern in DM/PM-IP patient and it found that DLCO. V brosis %,V hyper %, W hyper % and MLA total have signi cant statistical differences. On one hand, lower DLCO and higher V brosis % indicated that severity of brosis of NSIP/OP is worse than OP group. On the other hand,V hyper % and W hyper % are higher in the OP group demonstrate that − 901 to -1000 Hu area relatively increased under the in uence of non-aerated areas with consolidation pattern. Moreover, MLA total is an overall assessment of 100 in -1000 Hu.
Our study has some limitations. On one hand, it was a retrospective and single-center study of a small number of patients. One other hand, the diagnosis of NSIP, OP or NSIP/OP pattern based on histopathological diagnosis was carried out only in 8 patients.

Conclusion
QI of HRCT has good correlation with PFT parameters in ASS patient. NSIP/OP is a common pattern in ASS-IP, brosis severity of which is more severe than OP. Furthermore, QA can open a new sight on HRCT, as it is capable of provides accurate IP interpretation, especially the condition with lack pulmonary pathological evidence. And the high attenuation area between -100 to100 Hu should be paid attention to additional analysis except general threshold value range in these groups.

Declarations
Xueren Li is the guarantor of this research. Na Feng contributed to patient recruitment and patient followup. Shouchun Peng,Qi Wu were responsible for outcome measurement.Yuhua Zhang nished data acquisition. Huarui Zhang nished data analysis. All authors contributed to the drafting of this manuscript. All authors read and approved the nal manuscript.
Ethics approval and consent to participate Prior to its start, the study approved by the ethics committee of Pingjin Hospital, Tianjin, China. The reference number for the study is 2019-0007.

Consent for publication
Not applicable.