Virtual Library

Start Your Search

T. Koizumi



Author of

  • +

    P2.19 - Poster Session 2 - Imaging (ID 180)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Imaging, Staging & Screening
    • Presentations: 2
    • +

      P2.19-005 - Tumor growth rate of ground-glass nodules after long-term follow-up: The usefulness of a combination of initial computed tomography attenuation and tumor size as a predictor for tumor growth (ID 1023)

      09:30 - 09:30  |  Author(s): T. Koizumi

      • Abstract

      Background
      Pulmonary ground-glass nodules (GGNs) often grow over a long period of time. Elucidating the growth rate of GGNs is of great importance in deciding on follow-up intervals in clinical practice. The objective of this study was to clarify the natural history and tumor growth rate of GGNs after a long-term follow-up.

      Methods
      We retrospectively studied 121 cases of patients with GGNs, who had more than 50% of ground-glass opacity component and were monitored using high-resolution computed tomography (CT) for >2 years. Factors affecting the time to tumor progression were evaluated using the Kaplan–Meier method and Cox proportional hazard model analysis. Tumor growth rate was evaluated using specific growth rate (SGR), which was calculated using volume-doubling time (VDT) data (SGR = ln 2/VDT).

      Results
      During a median follow-up period of 59.7 months, GGNs showed enlargement or increase in attenuation in 66 patients and no change in 55 patients. In multivariate analysis using Cox proportional hazard model, the initial mean CT attenuation value > –670 Hounsfield unit (HU) and initial tumor size > 8 mm were independent predictive factors for time to tumor growth (p < 0.0001 and p = 0.0006, respectively). The Kaplan-Meier curves evaluating time to tumor progression, which was divided into 4 parts according to the initial mean CT attenuation value and initial tumor size, showed that the final estimated probability for patients with GGNs with a mean CT attenuation value < –670 HU and tumor size < 8 mm was 14.5%, whereas that for the other 3 groups was more than 50%, and with a significant difference (p < 0.0001). Comparing the tumor growth rates in these 4 groups, the SGR in GGNs with a mean CT attenuation value < –670 HU and tumor size < 8 mm was significantly lower than that in GGNs with a mean CT attenuation value > –670 HU and tumor size > 8 mm (p < 0.0001) and with a mean CT attenuation value > –670 HU and tumor size < 8 mm (p = 0.0005).

      Conclusion
      A combination of the measurement of the mean CT attenuation and the tumor size in the initial CT examination may be useful for predicting GGN growth. Lower-attenuation and smaller-size GGNs grow, but with slow growth rate, and, for patients with such GGNs, longer follow-up intervals may be useful in clinical practice.

    • +

      P2.19-006 - Novel computed tomography windows for predicting pathological invasion area in pulmonary subsolid tumors (ID 1158)

      09:30 - 09:30  |  Author(s): T. Koizumi

      • Abstract

      Background
      Preoperatively distinguishing non-invasive or minimally invasive lung adenocarcinomas from invasive lung adenocarcinomas on the basis of the classification proposed by the International Association for the Study of Lung Cancer, American Thoracic Society and European Respiratory Society (IASLC/ATS/ERS) is essential. However, in some cases, this distinction is potentially difficult to make by using the conventional computed tomography (CT) settings, i.e., lung and mediastinal windows. The objective of this study is to evaluate novel CT window settings for preoperative detection of pathologically invasive adenocarcinomas.

      Methods
      We retrospectively investigated 112 pathological T1a adenocarcinomas with a ground-glass opacity pattern ratio on preoperative CT of more than 50%. In the CT performed before an operation, the window level/window width was set as follows: -300/600, -250/500, -200/400, -150/300, and 30/400 (conventional mediastinal window). Under these viewing conditions, a maximum diameter of residual tumor was defined as the predictive invasion diameter. Receiver operating characteristic curve analyses were used to verify whether each viewing condition could be used to predict pathologically invasive adenocarcinomas that have a maximum invasion diameter of more than 5 mm.

      Results
      The area under the curve values for the -300/600, -250/500, -200/400, -150/300, and 30/400 settings were 0.87, 0.91, 0.86, 0.84, and 0.78, respectively. The sensitivity and specificity in the window setting of -250/500 was 98.1% and 67.8%, when the cut-off value for the predictive invasion diameter was set at 6 mm.

      Conclusion
      We determined novel CT window settings. The window level/window width setting of -250/500 might be useful for predicting pathological invasion of more than 5 mm in T1a lung adenocarcinomas.