To evaluate dose-volume relationships between the individual muscular and structural components of mastication their effect on trismus in locoregionally advanced head and neck cancer (HNC). This was a prospective, longitudinal study of 35 patients with stage III-IV HNC treated on the Accelerated Fractionation by Concomitant Boost (AFX-CB) arm of the Phase 2 EPIC Trial. All patients received 2 cycles of induction chemotherapy (Cetuximab 400 mg/m2 week 1; then 250 mg/m2, Paclitaxel 100 mg/m2 weeks 1-7, and Carboplatin AUC = 6 weeks 1 and 5) followed by the AFX-CB regimen as evaluated in RTOG 9003 and 0522 (72 Gy to primary tumor and involved nodes in 42 fractions and 54 Gy to elective neck in 30 fractions over 6 weeks) using intensity modulated radiation therapy with concurrent Cetuximab 250 mg/m2 (day 1, weekly x7) and Cisplatin 100mg/m2 (weeks 1,4). Patients were comprehensively evaluated for trismus using subjective and objective measures. Pre-therapy, all patients underwent prospective evaluation with CT scan of the head neck and detailed history and physical examination. Patients were re-evaluated with CT scan and physical examination every 3 to 4 months for the first two years followed by every 6 to 12 months thereafter. Dose to the pterygoid, masseter, and temporalis muscles and the mandibular condyle were calculated and correlated to outcome. Pre-therapy, no patients were found to have trismus. At 6 and 12 months, 14.3% and 20% of patients developed trismus, respectively. Statistically significant relationships between trismus and mean doses to the pterygoid, masseter, and temporalis muscles were observed at 6 and 12 months (p<0.05). Mean dose to the mandibular condyle was not significantly correlated to trismus at 6 months (p = 0.43) or 12 months (p = 0.25). We prospectively demonstrate that a dose-volume relationship exists when using AFX-CB. In order to minimize toxicity, efforts to reduce dose to these structures of mastication is essential.

Purpose: Current standard therapy for nasopharyngeal carcinoma (NPC) is concurrent chemoradiation based on randomized data. However, limited randomized data exist to support the addition of induction chemotherapy (ICT).Methods: 58 Patients with NPC were treated from 1990 to 2010. All patients received platinum-based ICT. All 58 patients were treated with chemoradiation, 57 in a week-on/week-off (WOWO) fashion. Concurrent chemotherapy included hydroxyurea/5-fluorouracil for all patients. Median radiation dose was 70 Gy. No patient received adjuvant chemotherapy.Results: AJCC 2009 stage was II = 13, III = 21, IVa = 13, and IVb = 11. Median follow-up for surviving patients was 66 months. Response to ICT was complete response (CR) 17% and partial response (PR) 64%. The CR rate after chemoradiation was 96%. Five-year actuarial freedom from local failure (FFLF), freedom from distant failure (FFDF), cause-specific survival (CSS), and overall survival (OS) was 98%, 90%, 90%, and 76%, respectively. Analysis of pediatric patients (n = 9) demonstrated 5-year actuarial FFLF, FFDF, CSS, and OS of 100%, 88%, 80%, and 80%, respectively.Conclusions: ICT followed by concurrent chemoradiation demonstrates excellent FFLF, FFDF, CSS, and OS with tolerable toxicity. Induction chemotherapy followed by concurrent chemoradiation for patients with NPC should be explored further in a randomized setting. (C) 2012 Elsevier Ltd. All rights reserved.

5519 Background: Cetuximab is commonly used in combination with either chemotherapy or radiation in the treatment of HNC, although the optimal way to integrate cetuximab with concurrent chemoradiation (CRT) remains unclear. We explored the addition of cetuximab to induction chemotherapy and two established CRT platforms in a randomized phase II trial. Methods: Patients with locoregionally advanced HNC were treated with cetuximab, carboplatin, paclitaxel induction chemotherapy for 2 cycles. Patients were randomized to A: cetuximab, 5-FU, hydroxyurea, and hyperfractionated week-on week-off RT (72-74Gy) (CetuxFHX), or B: cetuximab, cisplatin, accelerated radiation with concomitant boost (CetuxPX)(72Gy). Primary endpoints were 1- and 2-year progression-free (PFS), and overall survival (OS). Results: 110 patients with locoregionally advanced HNC (108 with Stage IV) were enrolled. 57.3% of patients had oropharyngeal (OP) primaries (A: 49%, B: 66%, p16 staining is pending). Induction response rate was 91.8%. 99.1% of patients developed a rash (≥grade 3 in 16.4%); ≥grade 3 neutropenia developed in 36.3% of patients. Overall 1- and 2-year survival rates were 98.3% and 89.5% in CetuxFHX, and 94.2% and 91.4% in CetuxPX arm, with 21.1months median follow-up (not statistically significant, p=0.27, logrank test). Progression free survival at 1 and 2 years was 86.0% and 82.3% in CetuxFHX, and 95.9% and 89.7% in CetuxPX (p=0.18). Grade ≥3 mucositis was present in 91.1%(A) and 94.3%(B) of patients; grade ≥3 dermatitis in 82.1% and 50.9% of patients. 95% of patients completed therapy. Treatment failures occurred in both OP and non-OP tumors. Conclusions: Cetuximab-based induction chemotherapy is well tolerated and active. Cetuximab can safely be integrated with both FHX and cisplatin based CRT with acceptable toxicities. Survival is favorable on both study arms suggesting that either platform could be investigated further. Data on HPV versus non-HPV-related tumors will be available.

6074 Background: Concurrent chemoradiotherapy (CRT) offers high functional organ preservation rates for locoregionally advanced head and neck cancer (LRAHNC) patients, but is associated with significant acute and chronic speech and swallowing toxicity. Recently, body mass index (BMI) has been suggested as a predictor of head and neck cancer patient outcome. In this analysis we sought to determine the impact of BMI on survival and toxicity outcomes in LRAHNC patients treated with CRT. Methods: 220 LRAHNC patients were treated on a multiinstitutional protocol consisting of induction carboplatin and paclitaxel followed by CRT. CRT was delivered for 4–5 cycles; each 14-day cycle consisted of 5 days concurrent paclitaxel, continuous infusion 5-FU, hydroxyurea, and 1.5 Gy twice daily radiation followed by 9 days without any treatment. Each patient's pre-treatment BMI was classified as overweight (BMI >= 25) or non-overweight (BMI < 25). As an independent variable, BMI was analyzed as a predictor of IndCT or CRT toxicity, locoregional control, and overall survival. BMI was analyzed as categorical variable, and also a continuous variable in a multivariate proportional hazards model. Results: There was no association between BMI and IndCT toxicity. During CRT overweight patients had significantly lower rates (24/103 vs 42/112) of grade 3 or higher neutropenia (p = 0.027), mucositis (p = 0.05), dermatitis (p = 0.028) and higher rates of anorexia (p = 0.05). Overweight patients had 12% long term PEG tube rate, compared to 34% of non-overweight patients (p < 0.001). On pooled survival analysis, patients with BMI > 25 had significantly better overall survival outcomes (mean 81.2 months, 95% CI 75.1–87.3 months) than patients with BMI < 25 (median 58.2 months; mean 56.5 months, 95% CI 49.6–63.3 months) (log-rank p < 0.001). Conclusions: Our data suggest patients with pre-treatment BMI > 25 experience lower rates of toxicity commonly associated with chemoradiation, and have a significantly better prognosis than patients with BMI < 25. Although the mechanism of BMI as an independent predictor of outcomes is unclear, we are continuing to explore mechanisms underlying this association. No significant financial relationships to disclose.

6067 Background: The prognosis for patients with locoregionally recurrent HNC remains poor. Re-irradiation with concomitant chemotherapy has demonstrated long-term tumor control rates of 20–30%, but a high percentage of pts recur within the irradiated field. Ad GV.EGR.TNF.11D is a second generation replication defective adenoviral vector that carries a human TNF-α gene linked to a chemotherapy and/or radiation inducible promoter [EGR-1]. TNF is a known radiation sensitizer; however, its systemic administration is associated with severe toxicities. Therefore, local administration of Ad GV.EGR.TNF.11D followed by chemoradiation should result in induction of regional TNF-α production and subsequent radiation enhancement without associated systemic side effects. Methods: In a phase I cohort dose escalation trial, eligible pts have locoregionally recurrent, previously radiated HNC and performance status 0–2. The dominant tumor site must be injectable. Patients receive 5 days of daily radiation at 200 cGy/fx with continuous infusion fluorouracil at 600 mg/m2/day x 5 and hydroxyurea at 500mg bid (FHX). Ad GV.EGR.TNF.11D is administered in escalating doses ranging from 4x109 to 4x1011 PU on day 2 of each cycle by direct intramural injection. 3 to 6 pts are entered on each cohort. Treatment cycles are repeated every other week. Results: 10 pts have been entered on three dose levels. Median age is 63 yrs (36–71); PS O: 2 pts; PS I: 6 pts; PS 2: 2 pts, prior surgery: 7 pts. Prior radiation: 60 to 75, Gy; 8 pts had received prior chemotherapy. Dose level 1 was expanded to 6 pts following observation of possibly related dose-limiting toxicity (carotid blow-out) in 1 pt. No subsequent dose- limiting toxicity was observed, and pts are currently entered on dose level 3. Overall, toxicities have consisted of grade III mucositis [8 pts] grade 2 or 3, dermatitis [5 and 4 pts], and are not suggestive of an interaction between Ad GV.EGR.TNF.11D and FHX. Following definition of the maximum tolerated dose level the dose of radiation will be escalated to 150 cGy/dx BID. Conclusions: Ad GV.EGR.TNF.11D can be added to concomitant FHX-based chemoradiotherapy. Updated toxicity and early efficacy data will be presented. No significant financial relationships to disclose.

To define outcome and prognostic factors in recurrent and second-primary head and neck cancer (HNC) treated with concomitant chemotherapy and re-irradiation (CRRT). 166 previously irradiated patients with recurrent or second primary squamous cell HNC (M0) were analyzed as a subset of 349 patients enrolled on 9 consecutive Phase I-II poor-prognosis protocols. Alternate week CRRT was delivered generally with 5-FU, hydroxyurea and a third agent (CDDP, CPT-11, gemcitabine, paclitaxel, or bevacuzumab). One protocol investigated induction chemotherapy prior to surgery or CRRT. Re-RT dose was 75 Gy for gross disease and 60-66 Gy after surgery. RT fractionation was 2 Gy QD or 1.5 Gy BID. Median follow-up for surviving pts was 52 mos (1-483 mos). 89% had recurrent disease; 11% had second primary. 81 patients (49%) underwent surgery prior to CRRT. 13% was previously treated with chemoradiotherapy. 56% received QD CRRT and 44% BID CRRT. Median re-RT dose was 66 Gy; median life-time RT dose was 131 Gy. Median overall survival (OS) and progression-free survival (PFS) were 10.3 mos and 6.5 mos, respectively. 3 year OS, DFS, loco-regional control (LRC), freedom from distant metastases (FFDM) were 19%, 18%, 48%, and 61%, respectively. On univariate analysis, surgery prior to CRRT, re-RT dose >58 Gy, BID fractionation, and lack of chemotherapy with previous RT were associated with improved OS. A trend for improved OS was observed with re-RT interval >36 mo and ECOG performance score ≤1 (p = 0.088 and 0.054, respectively). Poorly differentiated tumors were associated with worse FFDM (p = 0.038). No difference in any endpoint was seen between second primary and recurrent disease (3 year OS: 16% vs. 19%, p = 0.85). On multivariate analysis, surgery prior to CRRT, re-RT dose, and chemotherapy with prior RT were significant for OS, DFS, and LRC. BID fractionation improved LRC (p = 0.04). Tumor grade remained significant for FFDM (p = 0.009). CRRT was accompanied by substantial toxicity, including 18 treatment related deaths. There were 15 carotid blowouts (9%) and 17 cases of osteoradionecrosis (10.3%). There was one myelopathy (0.6%). For patients with recurrent or second primary HNC, undergoing CRRT, previous concurrent chemoradiotherapy predicts for worse outcome. BID re-RT is associated with improved LRC.

5558 Background: CTX regimens are often limited by severe acute toxicities and long-term functional impairment. This paper presents patient (pt) reported side effects and performance outcome in 3 sequential groups of stage IV HNC pts treated at the University of Chicago with IndCT and lowering radiation doses with the goal of maintaining survival while minimizing negative sequelae. Methods: Pts. were treated with IndCT followed by alternating weeks of concurrent CTX with lowering doses of radiotherapy to gross disease/ high risk microscopic/low risk microscopic disease: Group A - 75/60/45 Gy (n = 36 pts); Group B - 75/54/39 Gy (n = 34 pts); Group C - 72/51/36 Gy (n = 60 pts), respectively (treatment details & survival analyses presented separately). Pts were assessed for quality of life, performance (Performance Status Scale for HNC) and side effects (Radiotherapy Questionnaire) pretreatment, during CTX, at 3 & 12 months and annually thereafter. Controlling for baseline scores, differences between the groups were examined on- treatment and at 12-months. Results: Baseline and 12-month data were available for 81/114 pts. While all pts declined in performance (e.g., diet) and reported increased side effects (e.g., dry mouth, mouth pain, swallowing problems) on CTX, there were no statistically significant differences based on radiation dose. Similarly, there was improvement in most symptoms by 12 months with little difference among groups. In contrast, on normalcy of diet, fewer patients in Group C (20%) showed declines of more than 20 points from baseline compared to A (46%; p = .085) and B (45%, p = .09). Overall, 4% of patients were unable to take anything orally at 12-months with no difference among groups. Conclusions: This study showed no differences in pt reported symptoms on-treatment or at 12-months based on lowered radiation doses during CXT. On the other hand, while there continued to be some restriction in long-term diet, pts in the lowest radiation dose group showed somewhat less decline from baseline. These findings suggest the continued need to explore ways of decreasing both acute and long term toxicities in the use of CTX regimens. [Table: see text]

Recurrent head and neck cancer following definitive RT is almost uniformly fatal. Concurrent chemotherapy and re-irradiation (CRT) can cure some of these pts. In an attempt to identify prognostic factors for cure we analyzed our experience with concurrent chemotherapy and re-irradiation in 115 HNC pts. 350 pts were treated at our institution on a series of consecutive phase I/II HNC CRT protocols. We identified 115 pts with non-metastatic squamous cell HNC who failed prior RT. Initially, 14 pts were treated on a phase I dose finding HU, 5-FU, and RT(HFX) study. Further trials included novel agents. 23 pts were treated with CDDP/HFX, 42 with paclitaxel/HFX, 26 with gemcitabine (gem)/paclitaxel FX, and 10 with CPT-11/HFX. All pts underwent maximal surgical debulking if possible followed by alternate week CRT. We sought to deliver 70–75 Gy CRT, however, completely resected pts received less. To intensify RT, 71 pts received 1.5 Gy BID. Variables were analyzed on univariate (UVA) and multivariate (MVA) analyses to determine their impact on outcome. Median f/u for surviving pts was 67 mo(19–159). Median overall (OS) and progression free (PFS) survival were 11 and 8 mo (0.2–159), respectively. Median lifetime RT dose was 131 Gy. 5 year OS, PFS, loco-regional control (LRC), and freedom from distant metastases (FFDM) were 16%, 24%, 51%, 61%, respectively. At last f/u, 17 pts (15%) were alive and NED. On UVA, resection prior to CRT, complete response (CR), performance status (PS), and re-irradiation (re-RT) dose were associated with improved OS, PFS, and LRC. A trend for improved OS, PFS, and LRC was seen on UVA in pts with <3 curative therapy attempts (p = 0.06). Exclusion of pts who did not complete treatment, and those who died on treatment did not alter PFS, OS, and LRC outcome. MVA for OS revealed re-RT dose, CR, and pre-CRT surgery were independently prognostic. MVA for PFS revealed pre-CRT surgery, CR, and gem/CDDP were predictors of PFS. On UVA for LRC, novel agents (CDDP, paclitaxel, gem, or CPT-11; p = 0.0032) were significantly associated with better outcome, but only CR to therapy was associated with improved LRC on MVA. A trend for better LRC with pre-CRT surgery was also seen (p = 0.066). FFDM was improved with gem, CDDP, and paclitaxel chemotherapy on both MVA and UVA. A trend for improved FFDM was seen with CR to therapy (p = 0.055). Interestingly, increasing 5-FU dose intensity was associated with worse OS, PFS, and LRC on MVA. 66 non-operable pts were analyzed separately. 39% were alive at 1 yr and 19% were alive at 2 yrs. Median OS and PFS for non-resected pts was 9 and 7 mo, respectively vs. 17 and 10 mo for resected patients. In unresected pts, CR to CRT (p < 0.00001), and re-RT dose of > 58 Gy (P = .0207) were significantly associated with improved LRC, OS, and PFS. Volumetric tumor size in mL was analyzed as a continuous variable, and had no impact on LRC. 19 pts (17%) died of treatment related toxicity. 9 pts expired on treatment; 4 from sepsis, 3 from respiratory arrest, and 1 each from carotid blowout and pulmonary embolus. 57 pts were evaluable for long term swallowing toxicity. 14 (25%) were able to swallow a normal solid diet without difficulty. 7 (12%) pts were able to swallow solids for >1/2 of diet, and 8 (14%) pts had a primarily liquid diet. Prior to CRT 16 pts (28%) were g-tube dependent; an additional 12 pts (21%) became feeding tube dependent after. Of 41 pts with long term speech data, 3 (7%) had normal speech, 32 had intelligible speech (78%), 2 lost the ability to speak after re-RT (5%) and 4 (10%) were surgically mute. 1 pt with tumor next to the cord had myelopathy 53 mo after 130 Gy lifetime. 6 pts had carotid blowout. Concurrent CRT can be curative offering long term survival to a small % of pts. Due to substantial toxicity, full dose CRT should be performed in a protocol setting on pts with good PS. All pts should be treated with maximal surgical resection followed by high dose RT and chemotherapy

Purpose/Objective: Recurrent squamous cell carcinoma (SCCA) of the head and neck following radiotherapy (RT) has a poor prognosis with palliative chemotherapy offering 8–10 month median survival (MS). Current studies of re-irradiation (re-RT) in patients (pts) suggest that a small portion of these Pts may be curable despite the significant risk of toxicities. Based on our previous studies with re-irradiation, we sought to intensify both the chemotherapy and radiation schedules in a series of Phase I/II trials in the hope of improving outcome. We analyzed survival (OS), local regional tumor control (LRC), progression free survival (PFS), and toxicity following re-irradiation with concomitant chemotherapy (CRT) in pts with recurrent SCCA. Materials/Methods: Forty-one previously irradiated pts with recurrent SCCA but free of metastases were analyzed as a subset from two phase I/II chemoradiotherapy (CRT) trials. One cycle represents 5 days of concurrent chemoradiotherapy followed by a 9-day rest period. Chemotherapy consisted of oral hydroxyurea at 1000 mg given 30 minutes prior to RT, continuous infusion (CI) 5-fluorouracil at 800 mg/m2/day, and CI paclitaxel at 5–25 mg/m2/day or single infusion paclitaxel on day 1 of each cycle. RT was delivered in 2Gy fractions (11 pts) and later escalated to 1.5Gy BID (30 pts). Radiation portals encompassed gross disease and draining lymphatics. The cumulative lifetime spinal cord and brain stem doses did not exceed 50Gy and 54Gy, respectively. Fourteen pts underwent surgical salvage immediately prior to CRT (sCRT) while 27 received CRT alone. Pts were followed until recurrence or death. sCRT pts received 5 cycles of therapy (60 Gy) and CRT pts received 7 cycles (75 Gy). The median cumulative radiation dose was 134Gy. Results: Median follow-up for surviving pts is 72.5 (53–97) months. Four year MS of 11 months and OS of 22% was observed. MS and 4 year OS in the sCRT pts was 30.2 months and 43%, and in the CRT pts it was 10.2 months and 11% (p<0.007). LRC at 4 years was 54% for the entire cohort, 72% for sCRT, and 54% for CRT. There was no differences in local control (p<0.11). Four-year PFS was 57% in the sCRT group and 27% in the CRT group (p<0.05). Cox regression analysis revealed that radiation dose and surgery prior to re-irradiation were significant predictors of LRC, PFS, and OS. Grade 4/5 toxicity occurred in 32% of pts. Quality of life surveys from all 6 pts alive after 4 years demonstrated satisfaction with outcome despite the toxicities incurred. Conclusions: Concomitant chemotherapy and high-dose re-irradiation for recurrent head and neck cancer produces favorable PFS; although, acute and late treatment-related toxicity is significant. A small percentage of pts achieve long-term survival, with the more favorable group being those able to undergo at least partial surgical resection.