- Email: [email protected]
Outcomes of Stage I-II Squamous Cell Carcinoma of the Oral Tongue (OTSCC) Managed with Surgery Including an Elective Nodal Dissection: Potential Indications for Postoperative Radiation Therapy (PORT)
E386
International Journal of Radiation Oncology Biology Physics
after the median follow up of 46 months in IND group. The most common LNM were found in level II (41 in all 353 neck dissection, 11.6%), followed by level III (25/353, 7.1%), and level IV (4 in 219 neck dissection including level IV, 1.8%). The isolated level IV LNM was found in only 1 patients (0.5%). The poor histological differentiation was statistically significant risk factors of LNM (P < 0.008). Conclusion: Ipsilateral level II-IV irradiation is recommended in cN0 SSCC patients with pathological high and moderate differentiation. Contralateral level II-III irradiation may be considered in poor differentiation. However, the possible LNM side of the midline lesions need to be studied in future. Author Disclosure: Y. Zhang: None. J. Liu: None. J. Yi: None.
None. H. Paganetti: Research Grant; General Electric Global Research, National Cancer Institute. Honoraria; National Cancer Institute. Travel Expenses; National Cancer Institute. H. Lu: None. A.W. Chan: None.
2917 Temporal Lobe Necrosis After Proton for Nasopharyngeal Carcinoma: Predictive Factors and Clinical RBE Estimation Y. Zhang,1,2 W. Huo,2 J.A. Adams,2 N.N. Sanford,2,3 M.B. Lam,2 Y. Lu,2 S.I. Goldberg,2 H. Paganetti,2 H.M. Lu,2 and A.W. Chan2; 1Department of Oncology, Xiangya Hospital of Central South University, Changsha, China, 2Massachusetts General Hospital- Harvard Medical School, Radiation Oncology, Boston, MA, 3Harvard Radiation Oncology Program, Boston, MA Purpose/Objective(s): Temporal lobe necrosis (TLN) is one of the most debilitating radiotherapy complications that can impact the quality of life and survival of nasopharyngeal carcinoma (NPC) patients. We investigated the dose tolerance of the temporal lobe by examining various clinical and dosimetric parameters of TLN as well as dose differences between MonteCarlo (MC) simulation and the clinical treatment plan in regions of necrosis. Materials/Methods: Medical records and dose-volume data of 75 patients treated for newly diagnosed non-metastatic NPC between 1997 and 2013 at our institution were retrospectively reviewed and analyzed. The T-stage distribution was: T1 11%, T2 11%, T3 22%, T4 56%. Eighty-nine percent of patients received concurrent chemotherapy. Sixty-one patients were treated with proton radiotherapy and 14 were treated with intensity modulated radiotherapy (IMRT). The median dose to the primary GTV was 70Gy(RBE). For proton patients, the DVH of each temporal lobe (TL) was analyzed. Proton dose distributions were simulated using Monte Carlo method and compared with those obtained from the clinical treatment planning system. The relative biological equivalent (RBE) for protons was determined by using the reported TL tolerance with photons in NPC patients (D1cc Z 62.8Gy and D0.5cc Z 69Gy). Common Terminology Criteria for Adverse Events (CTCAE) Version 4.0 was used to grade TLN. Results: With a median follow-up of 5.6 years, one IMRT (7%) and 9 proton patients (14.8%) developed TLN, respectively, at a median time of 34 months (range: 9 e 82). Seventy-eight percent of patients with TLN had T3-4 disease. There were 4 grade 1, 4 grade 2, and 1 grade 3 cases of TLN. There was no grade 4 or 5 TLN. The 2- and 5-year rates of developing TLN after proton beam therapy was 3.6% and 14.4%, respectively. Asian race (pZ0.045) was the only clinical variable that was predictive for developing TLN. Age, gender, tumor volume, T-stage, diabetes, hypertension, cerebrovascular disease, smoking history, concurrent chemotherapy and induction chemotherapy were not statistically associated with TLN (p>0.05). The dose differences between the Monte Carlo simulation and clinical treatment plans were significant at the 10-40 Gy(RBE) and 70Gy(RBE) regions (p<0.05) but not at the 50-60Gy(RBE) regions (p>0.05). V10 to V67.5Gy(RBE), D0.5cc-D3cc, and Dmean were associated with TLN development (p<0.05). RBE for protons was estimated to range from 1.12 to 1.25, depending on the TL tolerances in the literature. Conclusion: Grade 3 or higher TLN is uncommon after proton beam radiotherapy for NPC. Asian race is associated with increased risk of TLN. The RBE of 1.1 that is currently used in clinical treatment is an underestimate. Author Disclosure: Y. Zhang: None. W. Huo: None. J.A. Adams: None. N.N. Sanford: None. M.B. Lam: None. Y. Lu: None. S.I. Goldberg:
2918 The Clinical Characteristics of Secondary Primary Tumors in Patients With Nasopharyngeal Carcinoma after Intensity-Modulated Radiation Therapy: A Retrospective Analysis W. Zhao,1 X. Zhu,2 and L. Li3; 1Department of Radiation Oncology, Cancer Hospital of Guangxi Medical University, nanning, China, 2Key Laboratory of High-Incidence-Tumor Prevention and Treatment, Guangxi Medical University, Nanning, Guang Xi, China, 3Department of Radiation Oncology, Cancer Hospital of Guangxi Medical University, nanning, China Purpose/Objective(s): To investigate the clinical characteristics associated with the risks of developing a secondary primary tumors (SPTs) in patients with nasopharyngeal carcinoma (NPC) who underwent intensitymodulated radiotherapy (IMRT). Materials/Methods: Data from 527 patients with biopsy-proven nonmetastatic NPC, who were treated with IMRT between January 2007 and December 2011, were analyzed retrospectively. The cumulative incidence of SPTs after IMRT completion was estimated using the KaplaneMeier method. Differences in cumulative incidence between groups were determined using the log-rank test. Results: The median follow-up time was 45.5 months (range, 4e97 months). Of 527 patients, 12 (2.3%) developed posttreatment SPTs (9 men and 3 women), 6 of which were located in the irradiated field. SPTs were mostly located in the upper aerodigestive tract (n Z 7), the head and neck (n Z 6), the lungs (n Z 3), and the tongue (n Z 2). The 1-, 3-, and 5-year cumulative SPT risk rates were 0.4%, 1.4%, and 3.1%, respectively, and the mean annual growth of cumulative incidence was w0.6%. The 1-, 3-, and 5-year cumulative in-field SPT risk rates were 0.4%, 0.8% and 1.5%, respectively, and the mean annual growth of in-field cumulative incidence was w0.3%. Univariate analysis revealed that sex, age, clinical stage, administration of chemotherapy, and overall IMRT duration did not significantly affect SPT risk. Conclusion: The 5-year SPTs incidence in the NPC patients after IMRT is concordant with or lower than that in previous 2D-RT studies .In patients with NPC who underwent IMRT, the upper aerodigestive tract was the most common site of SPTs, with lung cancer being the most common pathology. Further large-scale studies with longer-term follow-up are needed to determine risk factors associated with SPTs development after IMRT. Author Disclosure: W. Zhao: None. X. Zhu: None. L. Li: None.
2919 Outcomes of Stage I-II Squamous Cell Carcinoma of the Oral Tongue (OTSCC) Managed with Surgery Including an Elective Nodal Dissection: Potential Indications for Postoperative Radiation Therapy (PORT) M. Zhi1 and S. Iganej2; 1Kaiser Permanente Los Angeles Medical Center Los Angeles, CA, 2Southern California Permanente Medical Group, Los Angeles, CA Purpose/Objective(s): Patients with pathologic T1-2N0 oral tongue squamous cell carcinoma (OTSCC) are considered “low risk” and typically do not receive adjuvant treatment. We sought to characterize the outcomes of patients managed with surgery alone, their recurrence patterns, and any prognostics factors predictive of recurrence and/or survival. Materials/Methods: We retrospectively reviewed the records of 84 consecutive patients with newly-diagnosed, early stage pT1-2N0 OTSCC who underwent partial glossectomy and ipsilateral elective neck dissection from 2007-2013 at our institution. Patients who received PORT were excluded as well as those who had positive margins, pathological nodal involvement, or a history of prior or synchronous head and neck primaries.
Volume 99 Number 2S Supplement 2017 Results: The cohort was 60% male and was comprised of 62% pT1 and 65% grade 2-3 tumors. 51% of patients were current or former smokers. Median age was 58.5 (range: 20-87 years). With a median follow-up of 58 months (range: 3-131 months), 13 (16%) patients developed a relapse. The site of first relapse was isolated local in 5 patients, isolated regional in 2 patients, and combined locoregional in 6 patients. No patients developed a distant recurrence, either isolated or combined, at the time of first relapse. Regional recurrences were ipsilateral in 75% and contralateral in 25% of patients. Overall, the 5-year rates for local control (LC), regional control (RC), locoregional control (LRC), disease-free survival (DFS), diseasespecific survival (DSS), and overall survival (OS) were 89%, 91%, 86%, 80%, 93%, 87%, respectively. Pathologic T2 status was a predictor for worse outcomes across all endpoints. Perineural invasion (PNI) was a predictor for worse RC (pZ0.04), DSS (pZ0.001), and OS (pZ0.001). Margin 2mm was a predictor for worse LC (pZ0.0001) and PFS (pZ0.0011). Nine of the 49 (18%) patients with depth of invasion (DOI) 4mm suffered locoregional relapses (LRR). Of the 5 patients with pT2 and PNI, 2 (40%) developed isolated regional relapses. Of the 24 patients with pT2 and DOI 4mm, 5 (21%) had regional failures. Patients who developed neck recurrences experienced a significantly worse DSS (5year: 38% vs 100%; p<0.0001) and OS (5-year: 38% vs 92%; p<0.0001) compared to those who did not. All relapses underwent further salvage treatment; 8 (62%) patients were successfully salvaged with no evidence of disease at last follow-up. Conclusion: At our institution, patients with early stage, pT1-2N0 OTSCC exhibited relatively low rates of LRR and good overall prognosis. Regional relapse, however, significantly impacts OS adversely. Patients with pT2 disease who have PNI and/or DOI 4mm appear to be at considerable risk of regional relapse and should be counseled regarding PORT. When feasible, wider margins >2mm should be obtained surgically. Author Disclosure: M. Zhi: None. S. Iganej: None.
2920 The Risk Without Adaptive Replanning for HPV-Associated N2b Oropharyngeal Squamous Cell Carcinoma in Response to Anatomic Changes During Radiation Therapy H.J. Zhu, R.J. Amdur, B. Lu, C. Liu, C.E. Mercado, A.L. Holtzman, S.R.W. Nurkic, and W.M. Mendenhall; Department of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL Purpose/Objective(s): Replanning for head and neck cancer due to weight loss or tumor response takes substantial time, effort, and resources. There is no consensus on if, when, or how replanning should be done. Most published studies are limited by heterogeneity of the study population, variation in timing of replanning, and lack of clinically important dosimetric endpoints. The present study prospectively evaluated dosimetric changes at a uniform timepoint in a uniform patient population. Materials/Methods: The prospective program enrolled 10 consecutive patients with biopsy-proven p16-positive oropharyngeal squamous cell carinoma (OPX SCC) of clinical stage T1-3 and N2b who underwent primary radiotherapy and concurrent weekly chemotherapy between May and Dec 2016. All patients underwent repeat CT simulation on treatment day 21. Normal and target structures were contoured by the first author on both initial and second simulation CT. The target volumes and normal tissue constraints were defined per recent RTOG protocols. High-risk and standard-risk planning target volumes (PTV) were given 70Gy and 56Gy in 35 treatment days, respectively, using simultaneously integrated boost intensity-modulated radiotherapy (IMRT). We evaluated the effect of changes in anatomy without replanning by deformable registration of the initial dose map to the day 21 scan, which was scaled proportionally to its actual respective days of use. A combined dose map was then created with accumulated dose accounting for the anatomy change in the last 14 days of a 35-day treatment program. The impact of tumor response and weight change was analyzed in multivariate linear regression models. Results: No replanning resulted in underdose of both high-risk and standard-risk target volumes: PTV70 coverage reduced by 16.6 8.4% and
Poster Viewing E387 PTV56 reduced by 10.6 4.9%. Failure to meet the main coverage requirement (D95% Z prescription dose) occurred in 90% of patients. Failure to meet PTV coverage goals was associated with volume shrinkage of adenopathy (p <0.001), and multi-level adenopathy (p Z 0.009), but not weight loss. No replanning led to 1% change in mean dose of the contralateral parotid and <3% change in maximum dose to the spinal cord and brainstem. Overall, no normal tissues (spinal cord, brainstem, both parotids, larynx, pharyngeal constrictors, oral cavity) had dosimetric change to a significant degree. Conclusion: This is the first study of the adaptive replanning issue in a homogeneous study population with standardized evaluation parameters. Our findings were dramatically different from most prior publications. The primary risk without replanning was the underdosing of tumor targets, not an increased dose to normal structures. Our results suggest that most patients with HPV-associated oropharyngeal cancer who present with N2b adenopathy would benefit from replanning for at least the last 2 weeks of radiotherapy. Author Disclosure: H.J. Zhu: None. R.J. Amdur: Partnership; RadOnc eLearning Center, Inc. ; ABR, ACGME, AJCO, JCO, PRO. B. Lu: None. C. Liu: None. C.E. Mercado: None. A.L. Holtzman: None. S.R. Nurkic: None. W.M. Mendenhall: Employee; University of Florida.
2921 The Prognostic Value of E-cadherin Expression in Oral Cancer Patients D.M. Zielecka-De˛bska,1,2 A. Hao,3 R. Matkowski,1,2 J. Kornafel,1 and J. Szelachowska1,2; 1Department of Oncology, Wroclaw Medical University, Wroclaw, Poland., Wroclaw, Poland, 2Lower Silesia Oncology Center, Wroclaw, Poland., Wroclaw, Poland, 3Department of Pathomorphology and Oncological Cytology, Wroclaw Medical University, Wroclaw, Poland Purpose/Objective(s): Surgery for oral cancer is one of the main treatment methods. Status of surgical margins and stage of the disease (according to. TNM) are primary prognostic factors determining the eligibility of patients for postoperative treatment. In spite of an advanced qualification to post-operative treatment and low-stage of the disease (stage I and II), in approximately 16-42% of patients the later loco-regional recurrence is observed. E-cadherin is a membrane protein, generally found in epithelial tissues. Loss of E-cadherin increases the invasiveness of cancer cells in tumorigenesis. The aim of the study is to evaluate prognostic significance of E-cadherin on oral cancer patients. Materials/Methods: The study included 129 patients, suffering from squamous cell carcinoma of mouth floor (C04) or anterior part of tongue (C02), treated in the Lower Silesian Oncology Center in the years 19962011. 39 (30%) patients were diagnosed in stage II, 37 (29%) in stage III and 53 (41%) in stage IVa, respectively. All patients underwent microscopically radical resection of the primary tumor with simultaneous resection of regional lymph nodes and postoperative radiotherapy. E-cadherin expression was evaluated using a immunoreactivity Remmele scale (IRS), the percentage of stained cells (%) and the intensity of the color reaction (INT). Results: In the univariate Cox analysis (Table 1), percentage of cells expressing E-cadherin was a significant prognostic factor. Higher percentage of cells expressing E-cadherin was associated with a twofold increase risk of relapse and death. However, neither the intensity of the color Abstract 2921; Table 1 The percentage of 5-year locoregional recurrencefree survival (LRFS), disease-free survival (DFS), disease specific survival (DSS) and overall survival (OS) of patients based on the expression of E-cadherin in cells of squamous cell carcinoma E-cadherin expression (%)
5 -years LRFS
5-years DFS
5 -years DSS
5 -years OS
0-25% 26-75% >75% p-value
100% 76% 67% 0,020
100% 71% 58% 0,0050
100% 73% 58% 0,0029
73% 57% 44% 0,010
International Journal of Radiation Oncology Biology Physics
after the median follow up of 46 months in IND group. The most common LNM were found in level II (41 in all 353 neck dissection, 11.6%), followed by level III (25/353, 7.1%), and level IV (4 in 219 neck dissection including level IV, 1.8%). The isolated level IV LNM was found in only 1 patients (0.5%). The poor histological differentiation was statistically significant risk factors of LNM (P < 0.008). Conclusion: Ipsilateral level II-IV irradiation is recommended in cN0 SSCC patients with pathological high and moderate differentiation. Contralateral level II-III irradiation may be considered in poor differentiation. However, the possible LNM side of the midline lesions need to be studied in future. Author Disclosure: Y. Zhang: None. J. Liu: None. J. Yi: None.
None. H. Paganetti: Research Grant; General Electric Global Research, National Cancer Institute. Honoraria; National Cancer Institute. Travel Expenses; National Cancer Institute. H. Lu: None. A.W. Chan: None.
2917 Temporal Lobe Necrosis After Proton for Nasopharyngeal Carcinoma: Predictive Factors and Clinical RBE Estimation Y. Zhang,1,2 W. Huo,2 J.A. Adams,2 N.N. Sanford,2,3 M.B. Lam,2 Y. Lu,2 S.I. Goldberg,2 H. Paganetti,2 H.M. Lu,2 and A.W. Chan2; 1Department of Oncology, Xiangya Hospital of Central South University, Changsha, China, 2Massachusetts General Hospital- Harvard Medical School, Radiation Oncology, Boston, MA, 3Harvard Radiation Oncology Program, Boston, MA Purpose/Objective(s): Temporal lobe necrosis (TLN) is one of the most debilitating radiotherapy complications that can impact the quality of life and survival of nasopharyngeal carcinoma (NPC) patients. We investigated the dose tolerance of the temporal lobe by examining various clinical and dosimetric parameters of TLN as well as dose differences between MonteCarlo (MC) simulation and the clinical treatment plan in regions of necrosis. Materials/Methods: Medical records and dose-volume data of 75 patients treated for newly diagnosed non-metastatic NPC between 1997 and 2013 at our institution were retrospectively reviewed and analyzed. The T-stage distribution was: T1 11%, T2 11%, T3 22%, T4 56%. Eighty-nine percent of patients received concurrent chemotherapy. Sixty-one patients were treated with proton radiotherapy and 14 were treated with intensity modulated radiotherapy (IMRT). The median dose to the primary GTV was 70Gy(RBE). For proton patients, the DVH of each temporal lobe (TL) was analyzed. Proton dose distributions were simulated using Monte Carlo method and compared with those obtained from the clinical treatment planning system. The relative biological equivalent (RBE) for protons was determined by using the reported TL tolerance with photons in NPC patients (D1cc Z 62.8Gy and D0.5cc Z 69Gy). Common Terminology Criteria for Adverse Events (CTCAE) Version 4.0 was used to grade TLN. Results: With a median follow-up of 5.6 years, one IMRT (7%) and 9 proton patients (14.8%) developed TLN, respectively, at a median time of 34 months (range: 9 e 82). Seventy-eight percent of patients with TLN had T3-4 disease. There were 4 grade 1, 4 grade 2, and 1 grade 3 cases of TLN. There was no grade 4 or 5 TLN. The 2- and 5-year rates of developing TLN after proton beam therapy was 3.6% and 14.4%, respectively. Asian race (pZ0.045) was the only clinical variable that was predictive for developing TLN. Age, gender, tumor volume, T-stage, diabetes, hypertension, cerebrovascular disease, smoking history, concurrent chemotherapy and induction chemotherapy were not statistically associated with TLN (p>0.05). The dose differences between the Monte Carlo simulation and clinical treatment plans were significant at the 10-40 Gy(RBE) and 70Gy(RBE) regions (p<0.05) but not at the 50-60Gy(RBE) regions (p>0.05). V10 to V67.5Gy(RBE), D0.5cc-D3cc, and Dmean were associated with TLN development (p<0.05). RBE for protons was estimated to range from 1.12 to 1.25, depending on the TL tolerances in the literature. Conclusion: Grade 3 or higher TLN is uncommon after proton beam radiotherapy for NPC. Asian race is associated with increased risk of TLN. The RBE of 1.1 that is currently used in clinical treatment is an underestimate. Author Disclosure: Y. Zhang: None. W. Huo: None. J.A. Adams: None. N.N. Sanford: None. M.B. Lam: None. Y. Lu: None. S.I. Goldberg:
2918 The Clinical Characteristics of Secondary Primary Tumors in Patients With Nasopharyngeal Carcinoma after Intensity-Modulated Radiation Therapy: A Retrospective Analysis W. Zhao,1 X. Zhu,2 and L. Li3; 1Department of Radiation Oncology, Cancer Hospital of Guangxi Medical University, nanning, China, 2Key Laboratory of High-Incidence-Tumor Prevention and Treatment, Guangxi Medical University, Nanning, Guang Xi, China, 3Department of Radiation Oncology, Cancer Hospital of Guangxi Medical University, nanning, China Purpose/Objective(s): To investigate the clinical characteristics associated with the risks of developing a secondary primary tumors (SPTs) in patients with nasopharyngeal carcinoma (NPC) who underwent intensitymodulated radiotherapy (IMRT). Materials/Methods: Data from 527 patients with biopsy-proven nonmetastatic NPC, who were treated with IMRT between January 2007 and December 2011, were analyzed retrospectively. The cumulative incidence of SPTs after IMRT completion was estimated using the KaplaneMeier method. Differences in cumulative incidence between groups were determined using the log-rank test. Results: The median follow-up time was 45.5 months (range, 4e97 months). Of 527 patients, 12 (2.3%) developed posttreatment SPTs (9 men and 3 women), 6 of which were located in the irradiated field. SPTs were mostly located in the upper aerodigestive tract (n Z 7), the head and neck (n Z 6), the lungs (n Z 3), and the tongue (n Z 2). The 1-, 3-, and 5-year cumulative SPT risk rates were 0.4%, 1.4%, and 3.1%, respectively, and the mean annual growth of cumulative incidence was w0.6%. The 1-, 3-, and 5-year cumulative in-field SPT risk rates were 0.4%, 0.8% and 1.5%, respectively, and the mean annual growth of in-field cumulative incidence was w0.3%. Univariate analysis revealed that sex, age, clinical stage, administration of chemotherapy, and overall IMRT duration did not significantly affect SPT risk. Conclusion: The 5-year SPTs incidence in the NPC patients after IMRT is concordant with or lower than that in previous 2D-RT studies .In patients with NPC who underwent IMRT, the upper aerodigestive tract was the most common site of SPTs, with lung cancer being the most common pathology. Further large-scale studies with longer-term follow-up are needed to determine risk factors associated with SPTs development after IMRT. Author Disclosure: W. Zhao: None. X. Zhu: None. L. Li: None.
2919 Outcomes of Stage I-II Squamous Cell Carcinoma of the Oral Tongue (OTSCC) Managed with Surgery Including an Elective Nodal Dissection: Potential Indications for Postoperative Radiation Therapy (PORT) M. Zhi1 and S. Iganej2; 1Kaiser Permanente Los Angeles Medical Center Los Angeles, CA, 2Southern California Permanente Medical Group, Los Angeles, CA Purpose/Objective(s): Patients with pathologic T1-2N0 oral tongue squamous cell carcinoma (OTSCC) are considered “low risk” and typically do not receive adjuvant treatment. We sought to characterize the outcomes of patients managed with surgery alone, their recurrence patterns, and any prognostics factors predictive of recurrence and/or survival. Materials/Methods: We retrospectively reviewed the records of 84 consecutive patients with newly-diagnosed, early stage pT1-2N0 OTSCC who underwent partial glossectomy and ipsilateral elective neck dissection from 2007-2013 at our institution. Patients who received PORT were excluded as well as those who had positive margins, pathological nodal involvement, or a history of prior or synchronous head and neck primaries.
Volume 99 Number 2S Supplement 2017 Results: The cohort was 60% male and was comprised of 62% pT1 and 65% grade 2-3 tumors. 51% of patients were current or former smokers. Median age was 58.5 (range: 20-87 years). With a median follow-up of 58 months (range: 3-131 months), 13 (16%) patients developed a relapse. The site of first relapse was isolated local in 5 patients, isolated regional in 2 patients, and combined locoregional in 6 patients. No patients developed a distant recurrence, either isolated or combined, at the time of first relapse. Regional recurrences were ipsilateral in 75% and contralateral in 25% of patients. Overall, the 5-year rates for local control (LC), regional control (RC), locoregional control (LRC), disease-free survival (DFS), diseasespecific survival (DSS), and overall survival (OS) were 89%, 91%, 86%, 80%, 93%, 87%, respectively. Pathologic T2 status was a predictor for worse outcomes across all endpoints. Perineural invasion (PNI) was a predictor for worse RC (pZ0.04), DSS (pZ0.001), and OS (pZ0.001). Margin 2mm was a predictor for worse LC (pZ0.0001) and PFS (pZ0.0011). Nine of the 49 (18%) patients with depth of invasion (DOI) 4mm suffered locoregional relapses (LRR). Of the 5 patients with pT2 and PNI, 2 (40%) developed isolated regional relapses. Of the 24 patients with pT2 and DOI 4mm, 5 (21%) had regional failures. Patients who developed neck recurrences experienced a significantly worse DSS (5year: 38% vs 100%; p<0.0001) and OS (5-year: 38% vs 92%; p<0.0001) compared to those who did not. All relapses underwent further salvage treatment; 8 (62%) patients were successfully salvaged with no evidence of disease at last follow-up. Conclusion: At our institution, patients with early stage, pT1-2N0 OTSCC exhibited relatively low rates of LRR and good overall prognosis. Regional relapse, however, significantly impacts OS adversely. Patients with pT2 disease who have PNI and/or DOI 4mm appear to be at considerable risk of regional relapse and should be counseled regarding PORT. When feasible, wider margins >2mm should be obtained surgically. Author Disclosure: M. Zhi: None. S. Iganej: None.
2920 The Risk Without Adaptive Replanning for HPV-Associated N2b Oropharyngeal Squamous Cell Carcinoma in Response to Anatomic Changes During Radiation Therapy H.J. Zhu, R.J. Amdur, B. Lu, C. Liu, C.E. Mercado, A.L. Holtzman, S.R.W. Nurkic, and W.M. Mendenhall; Department of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL Purpose/Objective(s): Replanning for head and neck cancer due to weight loss or tumor response takes substantial time, effort, and resources. There is no consensus on if, when, or how replanning should be done. Most published studies are limited by heterogeneity of the study population, variation in timing of replanning, and lack of clinically important dosimetric endpoints. The present study prospectively evaluated dosimetric changes at a uniform timepoint in a uniform patient population. Materials/Methods: The prospective program enrolled 10 consecutive patients with biopsy-proven p16-positive oropharyngeal squamous cell carinoma (OPX SCC) of clinical stage T1-3 and N2b who underwent primary radiotherapy and concurrent weekly chemotherapy between May and Dec 2016. All patients underwent repeat CT simulation on treatment day 21. Normal and target structures were contoured by the first author on both initial and second simulation CT. The target volumes and normal tissue constraints were defined per recent RTOG protocols. High-risk and standard-risk planning target volumes (PTV) were given 70Gy and 56Gy in 35 treatment days, respectively, using simultaneously integrated boost intensity-modulated radiotherapy (IMRT). We evaluated the effect of changes in anatomy without replanning by deformable registration of the initial dose map to the day 21 scan, which was scaled proportionally to its actual respective days of use. A combined dose map was then created with accumulated dose accounting for the anatomy change in the last 14 days of a 35-day treatment program. The impact of tumor response and weight change was analyzed in multivariate linear regression models. Results: No replanning resulted in underdose of both high-risk and standard-risk target volumes: PTV70 coverage reduced by 16.6 8.4% and
Poster Viewing E387 PTV56 reduced by 10.6 4.9%. Failure to meet the main coverage requirement (D95% Z prescription dose) occurred in 90% of patients. Failure to meet PTV coverage goals was associated with volume shrinkage of adenopathy (p <0.001), and multi-level adenopathy (p Z 0.009), but not weight loss. No replanning led to 1% change in mean dose of the contralateral parotid and <3% change in maximum dose to the spinal cord and brainstem. Overall, no normal tissues (spinal cord, brainstem, both parotids, larynx, pharyngeal constrictors, oral cavity) had dosimetric change to a significant degree. Conclusion: This is the first study of the adaptive replanning issue in a homogeneous study population with standardized evaluation parameters. Our findings were dramatically different from most prior publications. The primary risk without replanning was the underdosing of tumor targets, not an increased dose to normal structures. Our results suggest that most patients with HPV-associated oropharyngeal cancer who present with N2b adenopathy would benefit from replanning for at least the last 2 weeks of radiotherapy. Author Disclosure: H.J. Zhu: None. R.J. Amdur: Partnership; RadOnc eLearning Center, Inc. ; ABR, ACGME, AJCO, JCO, PRO. B. Lu: None. C. Liu: None. C.E. Mercado: None. A.L. Holtzman: None. S.R. Nurkic: None. W.M. Mendenhall: Employee; University of Florida.
2921 The Prognostic Value of E-cadherin Expression in Oral Cancer Patients D.M. Zielecka-De˛bska,1,2 A. Hao,3 R. Matkowski,1,2 J. Kornafel,1 and J. Szelachowska1,2; 1Department of Oncology, Wroclaw Medical University, Wroclaw, Poland., Wroclaw, Poland, 2Lower Silesia Oncology Center, Wroclaw, Poland., Wroclaw, Poland, 3Department of Pathomorphology and Oncological Cytology, Wroclaw Medical University, Wroclaw, Poland Purpose/Objective(s): Surgery for oral cancer is one of the main treatment methods. Status of surgical margins and stage of the disease (according to. TNM) are primary prognostic factors determining the eligibility of patients for postoperative treatment. In spite of an advanced qualification to post-operative treatment and low-stage of the disease (stage I and II), in approximately 16-42% of patients the later loco-regional recurrence is observed. E-cadherin is a membrane protein, generally found in epithelial tissues. Loss of E-cadherin increases the invasiveness of cancer cells in tumorigenesis. The aim of the study is to evaluate prognostic significance of E-cadherin on oral cancer patients. Materials/Methods: The study included 129 patients, suffering from squamous cell carcinoma of mouth floor (C04) or anterior part of tongue (C02), treated in the Lower Silesian Oncology Center in the years 19962011. 39 (30%) patients were diagnosed in stage II, 37 (29%) in stage III and 53 (41%) in stage IVa, respectively. All patients underwent microscopically radical resection of the primary tumor with simultaneous resection of regional lymph nodes and postoperative radiotherapy. E-cadherin expression was evaluated using a immunoreactivity Remmele scale (IRS), the percentage of stained cells (%) and the intensity of the color reaction (INT). Results: In the univariate Cox analysis (Table 1), percentage of cells expressing E-cadherin was a significant prognostic factor. Higher percentage of cells expressing E-cadherin was associated with a twofold increase risk of relapse and death. However, neither the intensity of the color Abstract 2921; Table 1 The percentage of 5-year locoregional recurrencefree survival (LRFS), disease-free survival (DFS), disease specific survival (DSS) and overall survival (OS) of patients based on the expression of E-cadherin in cells of squamous cell carcinoma E-cadherin expression (%)
5 -years LRFS
5-years DFS
5 -years DSS
5 -years OS
0-25% 26-75% >75% p-value
100% 76% 67% 0,020
100% 71% 58% 0,0050
100% 73% 58% 0,0029
73% 57% 44% 0,010