Virtual Library

Start Your Search

A. Nelson



Author of

  • +

    P2.12 - Poster Session 2 - NSCLC Early Stage (ID 205)

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

      P2.12-015 - Investigation of the FDG PET-derived Total Glycolytic Activity (TGA) as Prognostic Tool in Patients with Early Stage Resected Non-Small Cell Lung Cancer (ID 1989)

      09:30 - 09:30  |  Author(s): A. Nelson

      • Abstract

      Background
      While the maximum standardized uptake values (SUV max) on the preoperative positron emission tomography (FDG PET-CT) have been associated with tumor recurrence in patients (pts) with resected early stage non-small cell lung cancer (ES-NSCLC), we investigated the prognostic role of tumor volume, using the Total Glycolytic Activity (TGA).

      Methods
      Pts with resected ES- NSCLC and single primary tumors without intrathoracic dissemination, staged with FDG PET-CT scans from 2006-2011 were included. Anonymized images of tumors were contoured with a commercially available semi-automatic gradient-based tool in order to derive the Metabolic Tumor Volume (MTV) and the SUV max/mean. The TGA (a product of the MTV and the SUV mean), was calculated. Patient-related, PET-derived and pathologic tumor characteristics were evaluated in univariable Cox proportional hazards models for association with Disease Free Survival, DFS. Factors significantly associated with DFS were included in multivariable models with either TGA or SUV max. Akaike Information Criterion (AIC) was used to compare the fit of the models. LogTGA was used due to a skewed distribution.

      Results
      170 pts with 121 PET scans were initially identified; 76 images were uploadable and 13 pts were excluded (1, small-cell lung cancer; 12, biopsy), leaving 63 analyzable pts. Median age was 69 (50-87) and 46% were males. Average time from FDG PET to surgery was 39 days (0-152). There were 55 (87 %) lobectomies/pneumonectomies and 8 (13%) wedge resections. Tumor histology was: 42 (66.7%) adenocarcinoma; 13 (20.6%) squamous cell carcinoma (SCCa); 4 (6.3%) adenosquamous (ASCa); 4 (6.3%) large cell (LCCa). Median tumor largest dimension was 2 cm (0.7-10.0); 12 (21.1%) tumors had lymphovascular invasion; 10 (18.5%) were node-positive (10 [17.2%] N1 and 2 [3.4%] N2); 3 pts had positive resection margin. Adjuvant treatment was given to 15 (24%) pts (13, chemotherapy; 4, radiation therapy). Median follow-up (FU) time was 32.2 months (mo) (0-83.8). Eighteen pts experienced disease progression and the first failure sites were: local (2); regional (4) and distant (16). Mean time to recurrence was 23.8 mo (median time, not reached) and 14 pts died. Median survival time (ST) was not reached; mean ST was 46.3 mo. Median MTV, median SUV max, median SUV mean and median TGA were as follows: 3.48 cc (range: 0.72-110.43); 6.25 (1.24-29.04); 3.56 (0.84-12.55); 10.22 (1.68-723.66). In univariable analysis, ASCa and LCCa were significantly (p = 0.011) associated with recurrence, compared to SCCa, and increasing logTGA showed a trend (p=0.12) for worse DFS. In multivariable analysis, log TGA and SUV max failed to reach statistical significance (p = 0.167 and 0.445, respectively); however, the log TGA model was found to fit the data slightly better than the SUV max model (AIC = 131.2 vs. 132.5, respectively). In the log TGA model, pts with ASCa and LCCa were 7.6 times more likely to have recurrence than those with SCCa (p = 0.04).

      Conclusion
      ASCa and LCCa histologies were associated with worse DFS . Log TGA may be a more informative measure for disease free survival than SUV max; however, further study of a larger size is needed.