EP-1312: Measurement of GTV delineation uncertainty for centrally recurrent gynaecological cancers

D. Bernstein, M. Llewelyn, A. Taylor, S. Nill, U. Oelfke
2016 Radiotherapy and Oncology  
ESTRO 35 2016 S615 ________________________________________________________________________________ Positron Emission Tomography/Computed Tomography (FDG-PET/CT) standardized uptake value (SUV) and total lesion glycolysis (TLG) with tumour characteristics and clinical response in a series of rectal cancer patients treated with neoadjuvant chemo-radiotherapy Material and Methods: Fifty-six patients were included in the present analysis. Pre-treatment PET maximum SUV (SUVmax), mean SUV and TLG of
more » ... mean SUV and TLG of primary tumour were calculated for each patient. The total dose of pelvic radiotherapy was 45-50.4 Gy, 1.8 Gy/fraction. Chemotherapy was delivered with capecitabina or 5-fluorouracil. Six to eight weeks after RT-CT, 44 patients (78.6%) had anterior rectal resection and 12 patients (21.4%) had abdominal pelvic resection (Miles). Tumor Regression Grade (TRG) (Mandard, 1994) was defined on surgical specimen. Complete regression (TRG1) was observed in 10/56 (17.9 %).The correlation between PET/CT results and histopathological data and tumour response was analyzed. Results: At the level of the primary tumour, SUVmax ranged from 4.17 to 54.06 (mean 22.46, median 18.89), SUV mean ranged from 6.22 to 32.64 (mean 13.42, median 11.09) and TLG ranged from 7.96 to3158.23 (mean 350.21, median 183.55).SUVmax (p= 0.05) and TLG (p= 0.002) significantly correlated with T-stage. Median SUVmax was significantly higher (p = 0.05) for lesions with partial response (PR, 46/56, 82.1%) than for lesions with complete response (CR, 34/54, 17.9%). Median TLG was significantly higher (p= 0.034) for lesions with partial response (PR, 45/54, 83.3%) than for lesions with complete response (CR, 9/54, 16.7%). SUVmean was not significantly correlated with T-stage (p= 0.074). Median SUVmean was higher for lesions with partial response (PR, 45/54, 83.3%) than for lesions with complete response (CR, 9/54, 16.7%) but without statistical significance (p =0.18). Conclusion: Our data suggest that pre-treatment FDG-PET/CT SUVmax and TLG are strongly associated with tumour primary tumour stage. Furthermore they correlate with prediction of tumour response after neoadjuvant treatment Electronic Poster: Clinical track: Gynaecological (endometrium, cervix, vagina, vulva) EP-1311 Chemoradiotherapy followed by surgery in patients with locally advanced cervical carcinoma Purpose or Objective: To evaluate pathological response and clinical outcomes in women with locally advanced cervical cancer treated with radiochemotherapy and surgery in a tertiary hospital. Material and Methods: In this retrospective analysis we have included 59 patients with cervical cancer (FIGO stages IB2-IVA) who were treated between December 2004 and July 2015 with concurrent chemoradiation therapy (CCRT) followed by surgery. The patients were treated with pelvic external beam radiotherapy at 46-50,4 Gy, 1,8-2 Gy/day. Based on CT or PET CT if aortic nodes were demonstrated, extended external beam radiotherapy was performed. We boosted nodes or parametria if they were affected (60-68 Gy, 2 Gy/day). After four weeks of treatment, patients received brachytherapy from 15 to 26 Gy in 3-6 fractions with 2D planification or 3D planification (n= 28), with a total tumour dose between 85 and 90 Gy. Concurrent chemotherapy with weekly platin and in some cases oral fluoropirimidine was administrated. Overall treatment time did not exceed 8 weeks. All patients completed surgery between 4-15 weeks after CCRT. Results: The median age was 52 years (range 30 and 77). Squamous cell carinoma was the most common subtype (81%). All patients received hysterectomy. 7 patients (12%) underwent lymphadenectomy. In global, 32 patients (54%) had a complete response, 20 (34%) a partial response and 7 (12 %) patients had residual microscopic disease in the pathologic analysis. With a median follow up of 53 months (range from 2 to 128 months) overall survival was 85% and disease free survival 81% Conclusion: Our results show that CCRT followed by surgery gets excellents outcomes with acceptable toxicity and may reduce local recurrences. Besides it enables assessment of the pathological response. Purpose or Objective: To quantify the magnitude of clinician uncertainty in GTV delineation for patients with recurrent gynaecological cancers. Material and Methods: GTV delineation uncertainty was retrospectively investigated in patients previously treated in our institution for centrally recurrent gynaecological cancer. In order to record clinician delineation uncertainty, clinicians were asked to draw 3 outlines per GTV; an inner GTV (GTV_I) corresponding to the innermost boundary the GTV is likely to have, an outer GTV (GTV_O)corresponding to the outmost boundary the GTV is likely to have and a clinical GTV (GTV_C) outlined in accordance with local treatment protocols. For GTV_C, each observer submitted a confidence score from 1 to 5, with 1 being no confidence in drawn and 5 complete confidence. For each patient, the 3 GTV's were delineated on a co-registered CT-MR, using a local rigid soft tissue registration, as well as on MR images only in order to identify how the co-registered CT information affects the decision process. Paired T Tests (p) were used to test for significance and Pearson correlation coefficient (r) for correlations. Results: To date, 18 recurrences from 17 patients were investigated by a single observer. For all 17 MR only contours and for 15 out of the 17 CT-MR contours, the GTV_O and GTV_C were identical. GTV_C ranged from 6.3 to 192.9cm3 for CT-MR contours and from 5.5 to 180.1cm3 for the MR only contours, with a mean ± standard deviation of 53.3±44.7cm3 and 39.3±40.4cm3 respectively. The reduction in GTV_I relative to GTV_C was 19.6±12.4cm3 (p<0.01) for CT-MR contours and 13.3±9.8cm3 (p<0.01) for MR only contours. For GTV_C, MR only contours were consistently smaller than CT-MR contours by 14.0±11.4cm3 (p<0.01). For GTV_I, MR only contours were smaller for 13 out of the 17 cases, with differences of 7.9±7.7cm3 (p<0.01). The 3D difference in the centre of mass (COM) between GTV_O and GTV_C was 2±2mm for the CT-MR contours and 2±1mm for the MR only contours. Scoring of GTV_C was significantly lower (p<0.01) for CT-MR contours relative to MR only contours, with scores of 2.8±0.6 and 3.6±0.7 respectively.
doi:10.1016/s0167-8140(16)32562-2 fatcat:fqdw6ru2rrhbji3h4ltepzqxue