Prognostic Parameters For Primary Tumor And Lymph Node Failure In Locally Advanced Nsclc Patients Treated With Concurrent Chemoradiation Therapy

Concurrent chemo-radiation therapy (CCRT) is treatment of choice in locally advanced non-small cell lung cancer (NSCLC) patients. Traditionally, the same radiation therapy (RT) dose is prescribed to the primary tumor (PT) and involved lymph nodes (LN), independent of volume and metabolic activity. However, volumes of PT and LN differ remarkably. The purpose of this study was to find prognostic parameters for locoregional control probability in locally advanced NSCLC patients. Between 2007 and 2011, 226 patients were treated with CCRT consisting of 24 fractions of 2.75 Gy intensity modulated RT with daily cisplatin (6 mg/m2) and included in this retrospective analysis. A staging FDG-PET-scan within 6 weeks prior to treatment was done. Follow-up (FU) consisted of a CT-thorax 4-6 weeks after treatment followed by X-thorax or CT-scans at 3-monthly intervals. PT and/or LN progression were classified based on FU scans and reviewed by two physicians. The volume of each PT and LN on the planning CT-scan and the SUVmax of the PET-scan were tested as prognostic factors for locoregional progression using Cox proportional hazard model with patient as random effect. Data were analyzed using R, package “coxme.” A total of 233 PTs (7 patients had no PT, 205 had 1 PT, 142 had 2 PTs) and 514 LNs (46 patients without LN, 46 had 1 LN, 44 had 2 LN and 90 had ≥3 LNs) were analyzed. The majority (92%) of the patients had stage III NSCLC and 78% had ≥N2. Median age was 64 years. At a median FU of 30 months (range 27-35) and median overall survival of 25 months (range 21-30), 35 PTs (15%) had progressed and 14 LNs (3%). Thirty-nine percent of the patients developed systemic disease. Average PT volume was 113.8 cc while the average LN volume was 13.6cc (p<0.001). Average SUVmax of PT and LN was 13.2, respectively 5.0 (p<0.001). The log(SUVmax) exhibited a linear relationship with log(volume), while the slopes were significantly different between PT and LN. Volume, SUVmax and lesion type (PT vs LN) were significant (p<0.001) as prognostic factor for progression in the univariate cox regression. In multivariate analysis, volume (p = 0.006) and lesion type (p = 0.004) remained significant. With a local recurrence of 15% and a regional recurrence of 3% in locally advanced NSCLC patients treated with CCRT, an excellent locoregional control is achieved with our regimen. The primary tumor was significantly associated with a higher risk of progression. A larger tumor volume and a higher SUVmax were associated as well with an increased risk. The strategy of de-escalating the RT dose to LN will be studied in future clinical trials.