Postoperative routine computed tomography allows detection of early-stage acute exacerbations of interstitial pneumonia

Authors:
  • KSKatsutoshi Seto
    Katsutoshi Seto
  • YIYuya Ishikawa
    Yuya Ishikawa
  • KKKoichiro Kimura
    Koichiro Kimura
  • TATakuya Adachi
    Takuya Adachi
  • AAAyaka Asakawa
    Ayaka Asakawa
  • MKMasashi Kobayashi
    Masashi Kobayashi
  • HIHironori Ishibashi
    Hironori Ishibashi
  • UTUkihide Tateishi
    Ukihide Tateishi

Abstract
Background: Postoperative acute exacerbation of interstitial pneumonia (IP-AE) is highly lethal after lung resection. Whether a symptom-free, routine computed tomography (rCT) can detect IP-AE early enough to justify pre-emptive steroids is unknown. We therefore aimed to assess the diagnostic accuracy of early postoperative rCT for detecting clinical IP-AE (cIP-AE) and to identify a high-specificity rCT score threshold that could support pre-emptive steroid treatment. Methods: This single-center, retrospective diagnostic-accuracy study analyzed 120 consecutive lung-cancer patients with underlying interstitial pneumonia (IP) who underwent high-resolution rCT within 7 days after surgery. Eligibility was based on preoperative high-resolution computed tomography (HRCT) demonstrating IP, irrespective of postoperative pathological confirmation. Two thoracic radiologists independently graded each scan on a five-point likelihood scale; a summed 10-point score (range, 2–10) was defined as the sum of the two readers’ five-point ratings, assigned retrospectively. Diagnostic accuracy was expressed as sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve (AUC). The ROC analysis assessed the ability of the rCT score to detect cIP-AE. For analysis, the proposed treatment threshold was the lowest score, giving 100% specificity. Results: IP-AE developed in 14 patients (11.6%). The ROC for cIP-AE detection yielded an AUC of 0.956. A summed score ≥9 achieved 100% specificity and detected 36% of presymptomatic events, whereas ≥6 increased sensitivity (86%) but lowered specificity (89%). Five patients received steroid pulse therapy solely on the basis of rCT; four of the five scans scored ≥9 and one scored 8. Thirty- and 90-day mortality were 0.8% and 3.3%, respectively, and no grade ≥3 steroid-related toxicity occurred. Conclusions: A single rCT obtained within 1 week of lung resection can reveal radiological IP-AE (rIP-AE) before clinical decline. Using a score ≥9 confines early steroid treatment to patients with near-certain disease, avoiding unnecessary exposure. Prospective studies should confirm whether this high-specificity strategy improves survival and resource utilization.

This publication has 33 references indexed in Scilit: