Virtual Library
Start Your Search
M. Rossi
Author of
-
+
MINI 18 - Radiation Topics in Localized NSCLC (ID 139)
- Event: WCLC 2015
- Type: Mini Oral
- Track: Treatment of Localized Disease - NSCLC
- Presentations: 1
-
+
MINI18.06 - Validation of High Risk Features on CT for Detection of Local Recurrence After SBRT for Stage I NSCLC (ID 2138)
16:45 - 18:15 | Author(s): M. Rossi
- Abstract
- Presentation
Background:
Fibrotic changes after SBRT for stage I NSCLC are difficult to distinguish from local recurrences (LR), hampering proper selection for salvage therapy. Huang et al. (1) defined CT high risk features (HRF) for detection of LR. This study attempts to validate these HRFs in an independent patient cohort.
Methods:
From a multicenter combined database of patients treated with SBRT for stage I NSCLC between 2006 and 2012, 53 LR were detected of which 14 were biopsy proven. The biopsy proven LR (N=14) were matched 1:2 to patients without LR (n=28) based on: 1) dose 2) PTV 3) follow up time 4) central/peripheral location 5) lung lobe. Of the resulting 42 patients 18 were male and 24 female with a median age of 73 years (range 56-89years). Median tumor size, PTV and dose were 2.3 cm (range 1.0-4.9cm), 49cc (range 9-166cc), 48 Gy (range 48-60Gy) in 4 fractions (range 3-8) respectively. Most tumors were peripheral (76%) and located in the upper lobes (55%). Median follow up (FU) was 36 months (range 14-78months) and median time to LR was 18 months (range 12-45months). For all patients, planning CT scans and at least two follow up scans were available. Two blinded observers scored eight HRFs for each scan. Sensitivity and specificity in predicting LR were assessed and compared using Fisher’s exact test. Analysis for best fit was done using AUC.
Results:
Results of sensitivity and specificity are shown in Table 1. The best performing HRF was cranio-caudal growth: sensitivity 86%, specificity 82%. The odds of LR increased on average by 2.6 (95%CI1.5-4.3) for each additional HRF detected, while the AUC was 0.86. The presence of ≥ 3 HRFs resulted in the best cut-off with sensitivity 79% and specificity 86%. Loss of linear margin and bulging margin were scored identical and therefore only the latter was included in the model. The two best combinations of HRFs were: 1) bulging margin & cranio-caudal growth, with a sensitivity of 93% and specificity of 82% or 2) bulging margin & enlarging opacity after 12 months, with a sensitivity of 86% and specificity of 89%. Table 1CT high risk factor for local recurrence Sensitivity (%) Specificity (%) p-value Any HRF 93 64 .001 enlarging opacity (≥5mm and ≥20%) 86 68 .003 sequential enlarging opacity 57 89 .002 enlarging opacity after 12 months 71 89 <.001 bulging margin 64 100 <.001 loss of linear margin 64 100 <.001 loss of air bronchograms 7 100 0.33 cranio-caudal growth (≥5mm and ≥20%) 86 82 <.001 new pleural effusion 14 93 0.59
Conclusion:
In this matched group of biopsy proven LR and controls, cranio-caudal growth was the best individual predictor of LR after SBRT. Combining HRF bulging margin with either cranio-caudal growth or enlarging opacity after 12 months resulted in higher sensitivities and specificities than number of HRFs. 1)Huang et al. Radiotherapy&Oncology 2013
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
