Virtual Library

Start Your Search

K. Frey



Author of

  • +

    P1.08 - Poster Session 1 - Radiotherapy (ID 195)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Radiation Oncology + Radiotherapy
    • Presentations: 1
    • +

      P1.08-026 - Functional dosimetric metrics based on the ventilation and perfusion SPECT-CT during the course of radiotherapy and association with radiation-induced lung toxicity in patients with non-small-cell lung cancer (ID 3325)

      09:30 - 16:30  |  Author(s): K. Frey

      • Abstract

      Background
      To investigate volume changes in single photon emission computed tomography (SPECT) ventilation (V) and perfusion (Q) - weighted functional lung (FL), and dose-volume histogram of functional lung (FDVH) from V or Q SPECT and factors associated with radiation-induced lung toxicity (RILT) during the course of radiotherapy (RT) in patients with non-small-cell lung cancer (NSCLC).

      Methods
      Seventy NSCLC patients treated with definitive RT were enrolled prospectively. V and Q SPECT-computed tomography (CT) were performed prior to and during RT at approximately 45 Gy. FL was created from the whole lung (WL) in V and Q SPECT using a threshold of 30% of the maximum uptake of normal lung. Patients had dose calculated on CT scan to provide a standard dose-volume histogram (DVH). V and Q SPECT scans provided FDVH and functional dosimetric parameters. From the FDVH; the percent of FL receiving from 5 to 60 Gy and the mean doses of functional lungs (FMLD) were computed. Differences of volume and volume change of WL and FL between pre-RT and during-RT were compared. Differences of functional and standard dosimetric parameters between the patients with and without RILT were compared. Clinical fibrosis and pneumonitis were graded according to Common Terminology Criteria for Adverse Events version 3.0, grade 2 and above were considered to be clinically significant.

      Results
      The accumulative incidence of ≥ grade 2 clinical fibrosis and pneumonitis were 23.4%. The volume of lung measured by using CT or V and Q SPECT did not change significantly during radiotherapy (All p>0.05). The difference of lung volume between standard CT or V and Q SPECT was not significant in patients with RILT and without RILT (All p>0.05). However, the volume of V and Q both functional lung changed -12.4% (95% CI, -29.6% to 4.9%) in patients with RILT and 13.3% (95% CI, 2.7% to 23.9%) in patients without RILT during RT (p=0.02). The difference of mean lung dose (MLD) measured by using CT or V and Q SPECT was not significant in patients with RILT and without RILT (All p>0.05). However, the MLD was significantly higher in patients with RILT in during-RT than pre-RT CT (17.1 Gy and 15.4 Gy, p=0.01); the FMLD was significantly higher in patients with RILT in during-RT than pre-RT V and Q SPECT (16.7 Gy and 13.0 Gy, p=0.03). The V20-V55 by using during-RT CT were higher in the patients with RILT than without RILT (V20:29.0% and 22.0%, p=0.06; V25: 26.4 % and 19.0 % p=0.04; V30: 23.7 % and 16.6 %, p=0.03; V35: 21.1 % and 14.6 %, p=0.03; V40: 18.9 % and 12.6 %, p=0.02; V45: 16.4 % and 10.9 %, p=0.03; V50: 14.3 % and 9.1 % p=0.03; V55: 11.6 % and 7.5 %, p=0.04). The V5-V60 by using during-RT V and Q SPECT was higher in patients with RILT than without RILT, but not statistically significant.

      Conclusion
      Functional DVH metrics of V and Q SPECT and during-RT CT based DVH may be more significant in their association with RILT than pre-CT. However, these results need to be validated.

  • +

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

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Imaging, Staging & Screening
    • Presentations: 1
    • +

      P2.19-012 - Inter-method Comparison of 18F-FDG PET Metabolic Response Assessment in Non-small Cell Lung Cancer (ID 2670)

      09:30 - 16:30  |  Author(s): K. Frey

      • Abstract

      Background
      This study aimed to: 1) compare qualitative and semi-quantitative assessment of [18]F-fluorodeoxyglucose PET based categorical metabolic response in patients with NSCLC and determine their value for prognosis prediction; 2) investigate the relationship between semi-quantitative assessments of post-treatment change of metabolic activity and survival and explore an optimal cutoff to distinguish a subset of responders with more favorable outcome.

      Methods
      This is a secondary analysis of prospective studies with IRB approval. Enrolled patients with NSCLC underwent PET/CT imaging within 2 weeks prior to (pre-RT PET/CT), at 4 weeks during and after radiation treatment (post-RT PET/CT). Post-RT metabolic therapeutic response was assessed using 1) visual assessment and 2) semi-quantitative measurement based on reduction in tumor FDG uptake; SUVmax normalized to mediastinal blood pool (NSUV-A). Interpretation of PET/CT scans was performed by three nuclear medicine physicians, blinded to clinical information. The three physicians did independent reads of the patients, performing a single read for each patient’s set of PET studies. Kappa coefficient was used to evaluate the agreement between categorical variables. Survival analysis and Cox proportional hazard regression model were adopted to analyze the effect of various response criteria on overall survival (OS) and progression free survival (PFS).

      Results
      Forty-four patients (36 male: 8 female) were eligible for analysis. The median interval between end of RT and post-RT PET/CT scan was 93 days. A poor agreement was observed between visual and semi-quantitative responses (Kappa coefficient = 0.393). Categorical responses were significantly correlated with both OS and PFS independent of employed response assessment criteria (either visual or semi-quantitative, p < 0.001) and patients with complete metabolic response (CMR) obtained the longest survival. As a continuous variable, reduction percentage of NSUV-A showed significant correlations with OS (hazard ratio, HR = 0.980, p < 0.001) and PFS (HR=0.984, p < 0.001). Analysis of OS and PFS consistently recommended NSUV-A reduction of sixty percent (60%) as another discriminative cutoff to distinguish patients with different outcome (p < 0.01).

      Conclusion
      There is a great discrepancy in metabolic response rates between qualitative and semi-quantitative methods. Categorical metabolic response criteria, either from qualitative visual assessment or a semi-quantitative method, demonstrate significant association with overall survival and progression free survival. Visual method offers a simpler approach that provides good information with regard to predicting OS and PFS, while the semi-quantitative method provides ordinal value that correlates with prolonged OS. As a continuum, the numerical percentage reduction in the normalized SUV is positively correlated with longer overall survival, reinforcing the prognostic value of metabolic change on FDG PET/CT. A sixty percent reduction of SUV may be the optimal cutoff for metabolic response to identify subsets of the PMR population with distinct outcomes, pending validation by an independent population.