Hideaki Yamakawa1, Takashi Ogura2, Shintaro Sato3, Tomotaka Nishizawa4, Rie Kawabe5, Tomohiro Oba6, Akari Kato7, Masanobu Horikoshi8, Keiichi Akasaka9, Masako Amano10, Kazuyoshi Kuwano11, Hiroki Sasaki12, Tomohisa Baba13, Hidekazu Matsushima14. 1. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan; Department of Respiratory Medicine, Tokyo Jikei University Hospital, 3-25-8 Nishi-shimbashi Minato-ku, Tokyo, 105-8461, Japan. Electronic address: hide1144@jikei.ac.jp. 2. Department of Respiratory Medicine, Kanagawa Cardiovascular and Respiratory Center, 6-16-1 Tomioka-higashi Kanazawa-ku, Yokohama, 236-0051, Japan. Electronic address: ogura@kanagawa-junko.jp. 3. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan. Electronic address: smallerss@hotmail.com. 4. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan. Electronic address: tn1230bungo@yahoo.co.jp. 5. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan. Electronic address: blackberry_wine1218@yahoo.co.jp. 6. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan. Electronic address: obatomohiro@hotmail.co.jp. 7. Department of Rheumatology, Saitama Red Cross Hospital, 1-5 Shintoshin Chuo-ku, Saitama, 330-8553, Japan. Electronic address: akarin110@hotmail.co.jp. 8. Department of Rheumatology, Saitama Red Cross Hospital, 1-5 Shintoshin Chuo-ku, Saitama, 330-8553, Japan. Electronic address: el_condor_vuelve@icloud.com. 9. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan. Electronic address: k-akasaka@med.niigata-u.ac. 10. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan. Electronic address: skyfield1021@gmail.com. 11. Department of Respiratory Medicine, Tokyo Jikei University Hospital, 3-25-8 Nishi-shimbashi Minato-ku, Tokyo, 105-8461, Japan. Electronic address: kkuwano@jikei.ac.jp. 12. Department of Radiology, Saitama Red Cross Hospital, 1-5 Shintoshin Chuo-ku, Saitama, 330-8553, Japan. Electronic address: sasaki.rad@gmail.com. 13. Department of Respiratory Medicine, Kanagawa Cardiovascular and Respiratory Center, 6-16-1 Tomioka-higashi Kanazawa-ku, Yokohama, 236-0051, Japan. Electronic address: baba@kanagawa-junko.jp. 14. Department of Respiratory Medicine, Saitama Red Cross Hospital, 1-5 Shintoshin, Chuo-ku, Saitama, 330-8553, Japan. Electronic address: himatsushima@sunny.ocn.ne.jp.
Abstract
BACKGROUND: Interstitial lung disease (ILD) is associated with high morbidity and mortality in patients with connective tissue disease (CTD). Because some patients with CTD overlap present with ILD first, with CTD diagnosed later, specific radiologic signs are needed to help differentiate each CTD or CTD-ILD from idiopathic ILD. OBJECTIVES: To determine whether specific CT findings can help differentiate CTD as rheumatoid arthritis (RA), systemic sclerosis (SSc), or polymyositis/dermatomyositis (PM/DM). METHODS: We analyzed 143 consecutive ILD patients with RA, SSc, or PM/DM. We assessed diagnostic accuracy of CT findings of CTD-ILD, CT pattern, and signs including "anterior upper lobe honeycomb-like lesion" and "low attenuation area (LAA) within an interstitial abnormality" for each CTD-ILD. Prognostic predictors were determined using Cox regression models. RESULTS: Subjects were 78 patients with RA-ILD, 38 with SSc-ILD, 24 with PM/DM-ILD, and 3 with overlapping CTD-ILD. High frequency of anterior upper lobe honeycomb-like lesion suggests that CTD-ILD is due to RA-ILD (22%) rather than SSc-ILD (8%) or PM/DM-ILD (8%), whereas LAA within an interstitial abnormality suggests that CTD-ILD is due to SSc-ILD (26%) rather than RA-ILD (4%) or PM/DM-ILD (0%). Multivariate analysis showed that while not associated with survival, current or ex-smoker, honeycombing, and acute exacerbation were negative prognostic factors of mortality. CONCLUSIONS: The tendency is high for RA-ILD, in which anterior upper lobe honeycomb-like lesion is a specific feature, to show UIP or NSIP/UIP pattern, combined emphysema, and honeycombing; SSc-ILD to show NSIP pattern and LAA within an interstitial abnormality; and PM/DM-ILD to show NSIP pattern and non-honeycombing.
BACKGROUND:Interstitial lung disease (ILD) is associated with high morbidity and mortality in patients with connective tissue disease (CTD). Because some patients with CTD overlap present with ILD first, with CTD diagnosed later, specific radiologic signs are needed to help differentiate each CTD or CTD-ILD from idiopathic ILD. OBJECTIVES: To determine whether specific CT findings can help differentiate CTD as rheumatoid arthritis (RA), systemic sclerosis (SSc), or polymyositis/dermatomyositis (PM/DM). METHODS: We analyzed 143 consecutive ILD patients with RA, SSc, or PM/DM. We assessed diagnostic accuracy of CT findings of CTD-ILD, CT pattern, and signs including "anterior upper lobe honeycomb-like lesion" and "low attenuation area (LAA) within an interstitial abnormality" for each CTD-ILD. Prognostic predictors were determined using Cox regression models. RESULTS: Subjects were 78 patients with RA-ILD, 38 with SSc-ILD, 24 with PM/DM-ILD, and 3 with overlapping CTD-ILD. High frequency of anterior upper lobe honeycomb-like lesion suggests that CTD-ILD is due to RA-ILD (22%) rather than SSc-ILD (8%) or PM/DM-ILD (8%), whereas LAA within an interstitial abnormality suggests that CTD-ILD is due to SSc-ILD (26%) rather than RA-ILD (4%) or PM/DM-ILD (0%). Multivariate analysis showed that while not associated with survival, current or ex-smoker, honeycombing, and acute exacerbation were negative prognostic factors of mortality. CONCLUSIONS: The tendency is high for RA-ILD, in which anterior upper lobe honeycomb-like lesion is a specific feature, to show UIP or NSIP/UIP pattern, combined emphysema, and honeycombing; SSc-ILD to show NSIP pattern and LAA within an interstitial abnormality; and PM/DM-ILD to show NSIP pattern and non-honeycombing.