- Email: [email protected]
PO-0663 DOWNSIZING AND DOWNSTAGING OF RECTAL CANCER AFTER SHORT COURSE RADIOTHERAPY FOLLOWED BY DELAYED SURGERY
S258
analyzed according to the criteria of the Japanese Research Society for Gastric Cancer by one radiologist. Results: Median follow-up was 56.3 months (range, 5.3-85.0 months). The 5-year regional failure free-survival (RFFS) rate was 63.6%. Regional failure (RF) as any component of first recurrence was 23.8% (91 patients), with isolated regional failure occurring in 49 patients (12.8%). Among 91 patients with RF, commonly involved recurrent lymph nodes were those in the No. 16b nodes (61.5%), No. 16a nodes (58.2%), No. 12 nodes (28.6%), No. 14 nodes (19.8%), No. 13 nodes (15.4%), and No. 9 nodes (15.4%). RFFS was adversely affected by advanced nodal stage (N3b vs. N3a) (hazard ratio, 2.02; 95% confidence interval, 1.40-2.91). The 5-year progression free-survival rate was 32.1% and overall survival 41.5%. Conclusions: The most prevalent nodal recurrence in patients with advanced gastric cancer was in the nodal basin outside D2 dissection field. Our findings may help physicians to construct lymph node target volume of gastric cancer after D2 dissection for radiation treatment. PO-0661 DOES SPLENECTOMY AFFECT THE OUTCOME OF POSTOPERATIVE RADIOCHEMOTHERAPY FOR GASTRIC CANCER? R. Suwinski1, J. Wydmanski1 1 Centre of Oncology - Institute MSC Gliwice, Radiation Oncology, Gliwice, Poland Purpose/Objective: The impact of splenectomy in the management of locally advanced gastric cancer remains controversial. An influence of splenectomy on tolerance and effectiveness of adjuvant postoperative radiochemotherapy has not been extensively studied. This created the basis for the present study. Materials and Methods: Between January 2001 and January 2006 351 patients had postoperative radiochemotherapy after surgery for gastric cancer. In general, prescription of radiochemotherapy was based on MacDonalds protocol, with loco-regional radiotherapy (45 Gy in 1.8 Gy per fraction over 5 weeks) and 5-Fu based chemotherapy (2 courses during radiotherapy, and 4 thereafter). Patients were in a good (ZUBROD 0) or fair (ZUBROD 1) performance (66% and 32% respectively). The highest proportion (71%) had pT3 tumors, while 12% and 17% had pT2 and pT3 tumors respectively. Only 78 patients (22%) appeared node negative. There were 254 males (72%) and 97 females (28%). Median age of the patients was 61 years (range 18-80 years). Weight loss of 10% or more was recorded in 173 patients (49%), while smaller loss in the remaining 178 patients (51%). Among 351 patients 120 (34%) had splenectomy, while 231 (66%) had not. The clinical characteristics of these two groups did not differ significantly. Survival curves were plotted using Kaplan-Meier method and compared using Cox f-test. Results: A statistical trend towards inferior survival after splenectomy was observed in a whole group of 351 patients (p=0.18, RR=1.22). Such trend was particularly apparent in a subset of patients with ≥10% weight loss (p=0.07, RR=1.41) and in a subset of patients younger than 70 years (p=0.12, RR=1.26). In 138 patients with both risk features (≥10% weight loss, age<70 years) splenectomy appeared to have a significant negative impact on overall survival (p=0.02, RR=1.59). The detrimental effect of splenectomy on survival resulted both from inferior loco-regional control and higher rate of distant metastases. Interestingly, the hematological tolerance of concurrent radiochemotherapy was better among the patients who had splenectomy, with 20% of patients with grade 1-2 leucopenia, compared to 47% of patients with grade 1-2 leucopenia among those with spleen preservation. Likewise, the average lymphocyte count at the end of chemotherapy was higher among the patients who had splenectomy, compared to those who did not have it (0.75 G/l vs. 0.30 G/l). Conclusions: While these results have limitations typical for the retrospective studies, the data suggest that splenectomy combined with postoperative radiochemotherapy for gastric cancer may result in inferior survival, particularly among young patients with ≥10% weight loss. PO-0662 PHASE II STUDY OF PREOPERATIVE HELICAL TOMOTHERAPY WITH A SIMULTANEOUS INTEGRATED BOOST FOR RECTAL CANCER A. Sermeus1, B. Engels2, K. Tournel2, H. Everaert3, A. Hoorens4, N. Christian2, G. Storme2, D. Verellen2, M. De Ridder2 1 Universitair Ziekenhuis Brussel, Gastro-Enterology, Brussels, Belgium 2 Universitair Ziekenhuis Brussel, Radiotherapy, Brussels, Belgium 3 Universitair Ziekenhuis Brussel, Nuclear Medicine, Brussels, Belgium 4 Universitair Ziekenhuis Brussel, Pathology, Brussels, Belgium
ESTRO 31
Purpose/Objective: The addition of concomitant chemotherapy to preoperative radiotherapy is considered the standard of care for patients with cT3-4 rectal cancer. The combined treatment modality increases the complete response rate and local control (LC), but has no impact on survival or the incidence of distant metastases. In addition, it is associated with considerable toxicity. As an alternative strategy, we explored prospectively, preoperative helical tomotherapy with a simultaneous integrated boost (SIB). Materials and Methods: A total of 108 patients were treated with intensity-modulated and image-guided radiotherapy using the Tomotherapy Hi-Art II system. A dose of 46 Gy, in daily fractions of 2 Gy, was delivered to the mesorectum and draining lymph nodes, without concomitant chemotherapy. Patients with an anticipated circumferential resection margin (CRM) of less than 2 mm, based on magnetic resonance imaging, received a SIB to the tumor up to a total dose of 55.2 Gy. Acute and late side effects were scored using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0. Results: A total of 102 patients presented with cT3-4 tumors; 57 patients entered the boost group and 51 the no-boost group. One patient in the no-boost group developed a radio-hypersensitivity reaction, resulting in a complete tumor remission, a Grade 3 acute and Grade 5 late enteritis. No other Grade ≥3 acute toxicities occurred. With a median follow-up of 32 months, Grade ≥3 late gastrointestinal and urinary toxicity were observed in 6% and 4% of the patients, respectively. The actuarial 2-year LC, progression-free survival and overall survival were 98%, 79%, and 93%. Conclusions: Preoperative helical tomotherapy displays a favorable acute toxicity profile in patients with cT3-4 rectal cancer. A SIB can be safely administered in patients with a narrow CRM and resulted in a promising LC. This strategy is currently being compared to standard chemoradiotherapy in multicentric phase III trial (number NCT01224392) PO-0663 DOWNSIZING AND DOWNSTAGING OF RECTAL CANCER AFTER SHORT COURSE RADIOTHERAPY FOLLOWED BY DELAYED SURGERY S. Faria1, N. Kopek1, T. Hijal1, S. Liberman2, P. Charlebois2, B. Stein2, S. Meterissian2, A. Meguerditchian2, E. Debroux3, D. Tataryn4 1 McGill University MGH, Radiation Oncology, Montréal, Canada 2 McGill University, Surgery, Montréal, Canada 3 St Luc Hospital, Surgery, Montréal, Canada 4 St Mary's Hospital, Surgery, Montréal, Canada Purpose/Objective: Two common approaches to neoadjuvant therapy in locally advanced, resectable rectal cancer are short-course (25Gy/5 fractions) alone followed by immediate surgery, and long-course (50.4Gy/28 fractions) combined with chemotherapy followed by delayed surgery. Phase III studies comparing these two schedules show similar outcome and toxicity, except for downsizing and down staging. Complete response was on the order of 1% versus 16% after short and long course respectively. We performed a prospective phase II study using the short course regimen followed by delayed surgery (to give time for the rectal cancer to respond) to assess downsizing and down staging Materials and Methods: 48 patients were treated between October 2008 and October 2011. All had adenocarcinoma of rectum (up to 14 cm from the anal verge). All patients were staged with at least MRI, CT scan and colonoscopy (often also with EUS and PET scan). All patients received a pelvic irradiation dose of 25 Gy in 5 fractions without chemotherapy, followed by surgery 6-8 weeks after. MRI preoperative stage and rectal cancer size were compared with pathological stage and size. Toxicity was assessed using the CTC v3 scoring system. Results: 46 patients were stage T3Nx before radiotherapy, one T2N1 and one T4N1 (Table). Treatment was well tolerated. When manifested, acute GI toxicity was pronounced during the 2nd week after radiation, mostly with diarrhea: 33 patients had grade 0-1, 12 grade 2, and 3 grade 3. Median interval between last radiotherapy and surgery was 52 days. Abdominoperineal resection was performed in 13 and low anterior resection in 35 patients. Median distance of tumor from anal verge was 8 cm. Median tumor size before surgery was: 4cm (1.9-7.3cm) and after surgery was: 2cm (0 – 6cm). Four patients had complete response, and 3 cases had only foci of disease after the surgery (7 complete clinical responses). 24/48 (50%) patients had downstaging. 46/48 (96%) patients had reduction of the primary cancer, and in more than half of these cases there was greater than 50% reduction. 7/48 (15%) patients had significant morbidity after
ESTRO 31
surgery: One patient died 2 days after of massive MI; infection = 3, DVT =1, rectal bleeding=1, peristomal hernia = 1 patient. Conclusions: Short course radiotherapy followed by delayed surgery is a practical and feasible treatment schedule. The regimen was generally well tolerated with acute toxicity typically occurring at the second week post-treatment. Significant downsizing was observed in the majority of patients with complete or near complete response achieved in 15% of patients. Further follow-up is necessary to assess long term toxicity.
S259
identical JCI values (investigators 3 and 4 in table). In 6/7(86%) of these the high MDC was due to errors of under-outlining. Table – JCI and MDC values for 7 investigators with a higher MDC than expected by JCI Investigator JCI MDC(mm) 1 0.59 111.62 2 0.56 92.10 3 0.54 47.42 4 0.54 145.84 5 0.49 546.34 6 0.67 37.60 7 0.71 11.6 Conclusions: This is the first analysis of variation of GTV outlining using the MDC tool in a clinical trial and has been shown to provide additional information than the JCI alone. Further work is needed to validate an acceptable value for MDC in a clinical setting and explore the use of a slice by slice MDC.
PO-0664 MEAN DISTANCE TO CONFORMITY AS A TOOL FOR ASSESSING OUTLINING IN THE UK SCOPE 1 OESOPHAGEAL CHEMORADIOTHERAPY TRIAL S. Gwynne1, C. Hurt2, E. Spezi3, T. Crosby4, J. Staffurth5, R. Jena6 1 Velindre NHS TRUST, Clinical Oncology, Cardiff Wales, United Kingdom 2 Cardiff University, Wales Cancer Trials Unit, Cardiff, United Kingdom 3 Velindre Cancer Centre, Medical Physics, Cardiff, United Kingdom 4 Velindre Cancer Centre, Clinical Oncology, Cardiff, United Kingdom 5 Cardiff University, Division of Medicine, Cardiff, United Kingdom 6 Cambridge University Hospitals NHS Foundation Trust, Oncology, Cambridge, United Kingdom Purpose/Objective: To assess outlining variation in the SCOPE 1 pretrial test case using the mean distance to conformity (MDC) tool. Materials and Methods: The National Cancer Research Institute SCOPE1 trial is a UK phase 3 RCT of dCRT with capecitabine and cisplatin with or without cetuximab. As part of the pre-trial QA programme a mid-oesophageal tumour test case was sent to each investigator. Investigator GTV (i-GTV) outlines were received from 50 investigators from 34 UK centres. Morphometric assessment of outlines against a pre-defined gs-GTV was undertaken using Mean Distance to Conformity (MDC) and Jaccard Conformity Index (JCI). The MDC is a shape based statistic that measures the mean displacement needed to transpose every voxel in the i-GTV onto gs-GTV. Whole volume assessment included MDC, MDC (O) and MDC (U), the latter two metrics describing over and under-outlining components of the MDC respectively. Based on theoretical modelling a MDC value of ≤2.2 mm was used to define acceptable conformity (corresponding to a JCI of ≥0.7). Results: The 50 i-GTVs had a median JCI of 0.69 (IQR: 0.62-0.70) and a median MDC of 2.29mm (IQR: 2.05-2.76). 12/50 (28%) had a JCI of ≥0.7 and 23/50(46%) had a MDC value of ≤2.2mm. 37/50(74%) and 42/50(84%) had a MDC (O) and MDC (U) of ≤2.2mm respectively. No outlining protocol deviations were seen in those with a MDC ≤2.2mm. The MDC component analysis indicated a greater magnitude of overoutlining than under-outlining -median MDC (O) 1.73mm (IQR: 1.212.15) vs median MDC (U) 0.40mm (IQR: 0.2-0.93). MDC is well correlated with JCI (Rho=-0.84). Anatomically, areas of over-outlining were inclusion of azygous vein, bronchial tissue, bronchus and pericardial sac/mediastinal fat and under-outlining a failure to include the whole oesophageal wall. MDC provided additional information than the JCI in this dataset. It identified 7 i-GTVs identified with greater discordance from the gs-GTV than would have been expected by the JCI alone (see table) and provided further information on the magnitude of discordance of two i-GTVs with
PO-0665 EVIDENT ESOPHAGEAL TUMOUR PROGRESSION BETWEEN THE DIAGNOSTIC FDG-PET AND THE FDG-PET FOR RADIOTHERAPY PLANNING C.T. Muijs1, J. Pruim2, J.C. Beukema1, M. Berveling1, J.T.H. Plukker3, J.A. Langendijk1 1 University Medical Center Groningen / University of Groningen, Radiation Oncology, Groningen, The Netherlands 2 University Medical Center Groningen / University of Groningen, Molecular imaging and nuclear medicine, Groningen, The Netherlands 3 University Medical Center Groningen / University of Groningen, Surgical Oncology, Groningen, The Netherlands Purpose/Objective: Esophageal tumour progression can be measured by an increased SUV, tumour length or nodal involvement and TNM stage on FDG-PET. In this study, we evaluated tumour progression within the time interval between the diagnostic FDG-PET and the FDG-PET-scan that was performed for radiotherapy treatment planning. Materials and Methods: FDG-PET(/CT)- scans were performed for staging purposes in patients with a primary esophageal tumour. In addition, the FDG-PET-scan was repeated in those patients elected for curative (chemo)radiation treatment, for radiotherapy treatment planning purposes, within 2 weeks after referral. All images were examined blinded by an experienced nuclear physician (JP). Subsequently, the two PET-scans of a single patient were only compared directly if there were any discrepancies between the initial blinded examinations. SUVmax, tumour length, lymph node metastases and distant metastases were scored. Results: The median time between the two PET scans was 22 days (range: 8-49). The SUVmax increased with more than 10% in 19 patients (42%), while it decreased in 11 patients (24%). Fourteen patients (31%) showed progression in tumour length with >1 cm and a reduction in length was found in 2 patients (4%). TNM-stage progression was found in twelve patients (27%). Occult mediastinal nodes were detected by the second PET in 8 patients (18%) and new distant metastases became manifest in 6 patients (13%). There were no significant prognostic factors for TNM progression. However, a trend was noted towards TNM progression for the type of PET-camera used (p=0.05) and for the interval between the two PET scans (p=0.09, 95%CI -0.9-12.5). In the patient group with TNM progression the interval between the 2 scans was on average 30 days vs. 24 days for the patients without TNM progression. Conclusions: Esophageal tumour progression in the relatively short interval between the diagnostic and the planning FDG-PET scan for radiotherapy was found in a substantial part of the patients. These results stress the importance of reducing the time for diagnostic procedures and preparation for radiotherapy treatment planning. PO-0666 PREDICTIVE FACTORS FOR PERICARDIAL EFFUSION IN OESO-PHAGEAL CANCER PATIENTS TREATED WITH CHEMORADIOTHERAPY K. Hayashi1, M. Nomura1, M. Kamata1, H. Kojima1, M. Kohzai1, K. Sumita1, S. Sawada1 1 Kansai Medical Univercity, Radiology, Hirakata, Japan
analyzed according to the criteria of the Japanese Research Society for Gastric Cancer by one radiologist. Results: Median follow-up was 56.3 months (range, 5.3-85.0 months). The 5-year regional failure free-survival (RFFS) rate was 63.6%. Regional failure (RF) as any component of first recurrence was 23.8% (91 patients), with isolated regional failure occurring in 49 patients (12.8%). Among 91 patients with RF, commonly involved recurrent lymph nodes were those in the No. 16b nodes (61.5%), No. 16a nodes (58.2%), No. 12 nodes (28.6%), No. 14 nodes (19.8%), No. 13 nodes (15.4%), and No. 9 nodes (15.4%). RFFS was adversely affected by advanced nodal stage (N3b vs. N3a) (hazard ratio, 2.02; 95% confidence interval, 1.40-2.91). The 5-year progression free-survival rate was 32.1% and overall survival 41.5%. Conclusions: The most prevalent nodal recurrence in patients with advanced gastric cancer was in the nodal basin outside D2 dissection field. Our findings may help physicians to construct lymph node target volume of gastric cancer after D2 dissection for radiation treatment. PO-0661 DOES SPLENECTOMY AFFECT THE OUTCOME OF POSTOPERATIVE RADIOCHEMOTHERAPY FOR GASTRIC CANCER? R. Suwinski1, J. Wydmanski1 1 Centre of Oncology - Institute MSC Gliwice, Radiation Oncology, Gliwice, Poland Purpose/Objective: The impact of splenectomy in the management of locally advanced gastric cancer remains controversial. An influence of splenectomy on tolerance and effectiveness of adjuvant postoperative radiochemotherapy has not been extensively studied. This created the basis for the present study. Materials and Methods: Between January 2001 and January 2006 351 patients had postoperative radiochemotherapy after surgery for gastric cancer. In general, prescription of radiochemotherapy was based on MacDonalds protocol, with loco-regional radiotherapy (45 Gy in 1.8 Gy per fraction over 5 weeks) and 5-Fu based chemotherapy (2 courses during radiotherapy, and 4 thereafter). Patients were in a good (ZUBROD 0) or fair (ZUBROD 1) performance (66% and 32% respectively). The highest proportion (71%) had pT3 tumors, while 12% and 17% had pT2 and pT3 tumors respectively. Only 78 patients (22%) appeared node negative. There were 254 males (72%) and 97 females (28%). Median age of the patients was 61 years (range 18-80 years). Weight loss of 10% or more was recorded in 173 patients (49%), while smaller loss in the remaining 178 patients (51%). Among 351 patients 120 (34%) had splenectomy, while 231 (66%) had not. The clinical characteristics of these two groups did not differ significantly. Survival curves were plotted using Kaplan-Meier method and compared using Cox f-test. Results: A statistical trend towards inferior survival after splenectomy was observed in a whole group of 351 patients (p=0.18, RR=1.22). Such trend was particularly apparent in a subset of patients with ≥10% weight loss (p=0.07, RR=1.41) and in a subset of patients younger than 70 years (p=0.12, RR=1.26). In 138 patients with both risk features (≥10% weight loss, age<70 years) splenectomy appeared to have a significant negative impact on overall survival (p=0.02, RR=1.59). The detrimental effect of splenectomy on survival resulted both from inferior loco-regional control and higher rate of distant metastases. Interestingly, the hematological tolerance of concurrent radiochemotherapy was better among the patients who had splenectomy, with 20% of patients with grade 1-2 leucopenia, compared to 47% of patients with grade 1-2 leucopenia among those with spleen preservation. Likewise, the average lymphocyte count at the end of chemotherapy was higher among the patients who had splenectomy, compared to those who did not have it (0.75 G/l vs. 0.30 G/l). Conclusions: While these results have limitations typical for the retrospective studies, the data suggest that splenectomy combined with postoperative radiochemotherapy for gastric cancer may result in inferior survival, particularly among young patients with ≥10% weight loss. PO-0662 PHASE II STUDY OF PREOPERATIVE HELICAL TOMOTHERAPY WITH A SIMULTANEOUS INTEGRATED BOOST FOR RECTAL CANCER A. Sermeus1, B. Engels2, K. Tournel2, H. Everaert3, A. Hoorens4, N. Christian2, G. Storme2, D. Verellen2, M. De Ridder2 1 Universitair Ziekenhuis Brussel, Gastro-Enterology, Brussels, Belgium 2 Universitair Ziekenhuis Brussel, Radiotherapy, Brussels, Belgium 3 Universitair Ziekenhuis Brussel, Nuclear Medicine, Brussels, Belgium 4 Universitair Ziekenhuis Brussel, Pathology, Brussels, Belgium
ESTRO 31
Purpose/Objective: The addition of concomitant chemotherapy to preoperative radiotherapy is considered the standard of care for patients with cT3-4 rectal cancer. The combined treatment modality increases the complete response rate and local control (LC), but has no impact on survival or the incidence of distant metastases. In addition, it is associated with considerable toxicity. As an alternative strategy, we explored prospectively, preoperative helical tomotherapy with a simultaneous integrated boost (SIB). Materials and Methods: A total of 108 patients were treated with intensity-modulated and image-guided radiotherapy using the Tomotherapy Hi-Art II system. A dose of 46 Gy, in daily fractions of 2 Gy, was delivered to the mesorectum and draining lymph nodes, without concomitant chemotherapy. Patients with an anticipated circumferential resection margin (CRM) of less than 2 mm, based on magnetic resonance imaging, received a SIB to the tumor up to a total dose of 55.2 Gy. Acute and late side effects were scored using the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0. Results: A total of 102 patients presented with cT3-4 tumors; 57 patients entered the boost group and 51 the no-boost group. One patient in the no-boost group developed a radio-hypersensitivity reaction, resulting in a complete tumor remission, a Grade 3 acute and Grade 5 late enteritis. No other Grade ≥3 acute toxicities occurred. With a median follow-up of 32 months, Grade ≥3 late gastrointestinal and urinary toxicity were observed in 6% and 4% of the patients, respectively. The actuarial 2-year LC, progression-free survival and overall survival were 98%, 79%, and 93%. Conclusions: Preoperative helical tomotherapy displays a favorable acute toxicity profile in patients with cT3-4 rectal cancer. A SIB can be safely administered in patients with a narrow CRM and resulted in a promising LC. This strategy is currently being compared to standard chemoradiotherapy in multicentric phase III trial (number NCT01224392) PO-0663 DOWNSIZING AND DOWNSTAGING OF RECTAL CANCER AFTER SHORT COURSE RADIOTHERAPY FOLLOWED BY DELAYED SURGERY S. Faria1, N. Kopek1, T. Hijal1, S. Liberman2, P. Charlebois2, B. Stein2, S. Meterissian2, A. Meguerditchian2, E. Debroux3, D. Tataryn4 1 McGill University MGH, Radiation Oncology, Montréal, Canada 2 McGill University, Surgery, Montréal, Canada 3 St Luc Hospital, Surgery, Montréal, Canada 4 St Mary's Hospital, Surgery, Montréal, Canada Purpose/Objective: Two common approaches to neoadjuvant therapy in locally advanced, resectable rectal cancer are short-course (25Gy/5 fractions) alone followed by immediate surgery, and long-course (50.4Gy/28 fractions) combined with chemotherapy followed by delayed surgery. Phase III studies comparing these two schedules show similar outcome and toxicity, except for downsizing and down staging. Complete response was on the order of 1% versus 16% after short and long course respectively. We performed a prospective phase II study using the short course regimen followed by delayed surgery (to give time for the rectal cancer to respond) to assess downsizing and down staging Materials and Methods: 48 patients were treated between October 2008 and October 2011. All had adenocarcinoma of rectum (up to 14 cm from the anal verge). All patients were staged with at least MRI, CT scan and colonoscopy (often also with EUS and PET scan). All patients received a pelvic irradiation dose of 25 Gy in 5 fractions without chemotherapy, followed by surgery 6-8 weeks after. MRI preoperative stage and rectal cancer size were compared with pathological stage and size. Toxicity was assessed using the CTC v3 scoring system. Results: 46 patients were stage T3Nx before radiotherapy, one T2N1 and one T4N1 (Table). Treatment was well tolerated. When manifested, acute GI toxicity was pronounced during the 2nd week after radiation, mostly with diarrhea: 33 patients had grade 0-1, 12 grade 2, and 3 grade 3. Median interval between last radiotherapy and surgery was 52 days. Abdominoperineal resection was performed in 13 and low anterior resection in 35 patients. Median distance of tumor from anal verge was 8 cm. Median tumor size before surgery was: 4cm (1.9-7.3cm) and after surgery was: 2cm (0 – 6cm). Four patients had complete response, and 3 cases had only foci of disease after the surgery (7 complete clinical responses). 24/48 (50%) patients had downstaging. 46/48 (96%) patients had reduction of the primary cancer, and in more than half of these cases there was greater than 50% reduction. 7/48 (15%) patients had significant morbidity after
ESTRO 31
surgery: One patient died 2 days after of massive MI; infection = 3, DVT =1, rectal bleeding=1, peristomal hernia = 1 patient. Conclusions: Short course radiotherapy followed by delayed surgery is a practical and feasible treatment schedule. The regimen was generally well tolerated with acute toxicity typically occurring at the second week post-treatment. Significant downsizing was observed in the majority of patients with complete or near complete response achieved in 15% of patients. Further follow-up is necessary to assess long term toxicity.
S259
identical JCI values (investigators 3 and 4 in table). In 6/7(86%) of these the high MDC was due to errors of under-outlining. Table – JCI and MDC values for 7 investigators with a higher MDC than expected by JCI Investigator JCI MDC(mm) 1 0.59 111.62 2 0.56 92.10 3 0.54 47.42 4 0.54 145.84 5 0.49 546.34 6 0.67 37.60 7 0.71 11.6 Conclusions: This is the first analysis of variation of GTV outlining using the MDC tool in a clinical trial and has been shown to provide additional information than the JCI alone. Further work is needed to validate an acceptable value for MDC in a clinical setting and explore the use of a slice by slice MDC.
PO-0664 MEAN DISTANCE TO CONFORMITY AS A TOOL FOR ASSESSING OUTLINING IN THE UK SCOPE 1 OESOPHAGEAL CHEMORADIOTHERAPY TRIAL S. Gwynne1, C. Hurt2, E. Spezi3, T. Crosby4, J. Staffurth5, R. Jena6 1 Velindre NHS TRUST, Clinical Oncology, Cardiff Wales, United Kingdom 2 Cardiff University, Wales Cancer Trials Unit, Cardiff, United Kingdom 3 Velindre Cancer Centre, Medical Physics, Cardiff, United Kingdom 4 Velindre Cancer Centre, Clinical Oncology, Cardiff, United Kingdom 5 Cardiff University, Division of Medicine, Cardiff, United Kingdom 6 Cambridge University Hospitals NHS Foundation Trust, Oncology, Cambridge, United Kingdom Purpose/Objective: To assess outlining variation in the SCOPE 1 pretrial test case using the mean distance to conformity (MDC) tool. Materials and Methods: The National Cancer Research Institute SCOPE1 trial is a UK phase 3 RCT of dCRT with capecitabine and cisplatin with or without cetuximab. As part of the pre-trial QA programme a mid-oesophageal tumour test case was sent to each investigator. Investigator GTV (i-GTV) outlines were received from 50 investigators from 34 UK centres. Morphometric assessment of outlines against a pre-defined gs-GTV was undertaken using Mean Distance to Conformity (MDC) and Jaccard Conformity Index (JCI). The MDC is a shape based statistic that measures the mean displacement needed to transpose every voxel in the i-GTV onto gs-GTV. Whole volume assessment included MDC, MDC (O) and MDC (U), the latter two metrics describing over and under-outlining components of the MDC respectively. Based on theoretical modelling a MDC value of ≤2.2 mm was used to define acceptable conformity (corresponding to a JCI of ≥0.7). Results: The 50 i-GTVs had a median JCI of 0.69 (IQR: 0.62-0.70) and a median MDC of 2.29mm (IQR: 2.05-2.76). 12/50 (28%) had a JCI of ≥0.7 and 23/50(46%) had a MDC value of ≤2.2mm. 37/50(74%) and 42/50(84%) had a MDC (O) and MDC (U) of ≤2.2mm respectively. No outlining protocol deviations were seen in those with a MDC ≤2.2mm. The MDC component analysis indicated a greater magnitude of overoutlining than under-outlining -median MDC (O) 1.73mm (IQR: 1.212.15) vs median MDC (U) 0.40mm (IQR: 0.2-0.93). MDC is well correlated with JCI (Rho=-0.84). Anatomically, areas of over-outlining were inclusion of azygous vein, bronchial tissue, bronchus and pericardial sac/mediastinal fat and under-outlining a failure to include the whole oesophageal wall. MDC provided additional information than the JCI in this dataset. It identified 7 i-GTVs identified with greater discordance from the gs-GTV than would have been expected by the JCI alone (see table) and provided further information on the magnitude of discordance of two i-GTVs with
PO-0665 EVIDENT ESOPHAGEAL TUMOUR PROGRESSION BETWEEN THE DIAGNOSTIC FDG-PET AND THE FDG-PET FOR RADIOTHERAPY PLANNING C.T. Muijs1, J. Pruim2, J.C. Beukema1, M. Berveling1, J.T.H. Plukker3, J.A. Langendijk1 1 University Medical Center Groningen / University of Groningen, Radiation Oncology, Groningen, The Netherlands 2 University Medical Center Groningen / University of Groningen, Molecular imaging and nuclear medicine, Groningen, The Netherlands 3 University Medical Center Groningen / University of Groningen, Surgical Oncology, Groningen, The Netherlands Purpose/Objective: Esophageal tumour progression can be measured by an increased SUV, tumour length or nodal involvement and TNM stage on FDG-PET. In this study, we evaluated tumour progression within the time interval between the diagnostic FDG-PET and the FDG-PET-scan that was performed for radiotherapy treatment planning. Materials and Methods: FDG-PET(/CT)- scans were performed for staging purposes in patients with a primary esophageal tumour. In addition, the FDG-PET-scan was repeated in those patients elected for curative (chemo)radiation treatment, for radiotherapy treatment planning purposes, within 2 weeks after referral. All images were examined blinded by an experienced nuclear physician (JP). Subsequently, the two PET-scans of a single patient were only compared directly if there were any discrepancies between the initial blinded examinations. SUVmax, tumour length, lymph node metastases and distant metastases were scored. Results: The median time between the two PET scans was 22 days (range: 8-49). The SUVmax increased with more than 10% in 19 patients (42%), while it decreased in 11 patients (24%). Fourteen patients (31%) showed progression in tumour length with >1 cm and a reduction in length was found in 2 patients (4%). TNM-stage progression was found in twelve patients (27%). Occult mediastinal nodes were detected by the second PET in 8 patients (18%) and new distant metastases became manifest in 6 patients (13%). There were no significant prognostic factors for TNM progression. However, a trend was noted towards TNM progression for the type of PET-camera used (p=0.05) and for the interval between the two PET scans (p=0.09, 95%CI -0.9-12.5). In the patient group with TNM progression the interval between the 2 scans was on average 30 days vs. 24 days for the patients without TNM progression. Conclusions: Esophageal tumour progression in the relatively short interval between the diagnostic and the planning FDG-PET scan for radiotherapy was found in a substantial part of the patients. These results stress the importance of reducing the time for diagnostic procedures and preparation for radiotherapy treatment planning. PO-0666 PREDICTIVE FACTORS FOR PERICARDIAL EFFUSION IN OESO-PHAGEAL CANCER PATIENTS TREATED WITH CHEMORADIOTHERAPY K. Hayashi1, M. Nomura1, M. Kamata1, H. Kojima1, M. Kohzai1, K. Sumita1, S. Sawada1 1 Kansai Medical Univercity, Radiology, Hirakata, Japan