Temporal Stability and Reproducibility of FDG-PET–Based Dose Painting Targets in Head and Neck Cancer

International Journal of Radiation Oncology  Biology  Physics

S514 Scientific Abstract 2767; Table All patients (n Z 151) HPV+ (n Z 86) HPV- (n Z 18)

3-year outcomes for cN+ OPC patients with rECE vs without rECE LRC

DC

PFS

CSS

OS

83% vs 96%, P Z .006 94% vs 100%, P Z .098 30% vs 74%, P Z .022

77% vs 99%, P<.001 87% vs 97%, P Z .098 19% vs 100%, P Z .002

66% vs 88%, P Z .001 85% vs 82%, P Z .712 0% vs 74%, P Z .001

78% vs 98%, P Z .001 90% vs 96%, P Z .203 44% vs 100%, P Z .036

69% vs 91%, P Z .003 88% vs 81%, P Z .854 13% vs 100%, P Z .001

Conclusions: Our analysis suggests that unilateral RT (+/- chemotherapy) for node-positive tonsillar cancer in the HPV era may be appropriate, in well-selected patients. This is notable because a significant number of patients on our study had N2 neck disease, which has historically been considered a contraindication to unilateral treatment. Despite the concern that HPV-positive pts are more likely to have bilateral neck disease, in this series we observed no contralateral failures. Unilateral treatment yielded good functional outcomes with no PEG-dependency, and low rates of xerostomia. These findings should be validated in a larger series. Author Disclosure: M.E. Gamez: None. K. Hu: None. J.N. Lukens: None. W.F. Mourad: None. M. Agarwal: None. R. Metcalfe-Klaw: None. T. Tran: None. M. Persky: None. A. Jacobson: None. M. Urken: None. Z. Li: None. B. Culliney: None. L.B. Harrison: None.

2767 Is Radiographic Extracapsular Extension Prognostic in Human PapillomaviruseRelated Oropharyngeal Cancers? J.T. Liu,1 B.H. Kann,1 B. De,1 M. Buckstein,1 R.L. Bakst,1 E.M. Genden,2 M.R. Posner,3 P.M. Som,4 and V. Gupta1; 1Department of Radiation Oncology, Icahn School of Medicine at Mount Sinai, New York City, NY, 2 Department of Otolaryngology, Icahn School of Medicine at Mount Sinai, New York City, NY, 3Department of Hematology and Medical Oncology, Icahn School of Medicine at Mount Sinai, New York City, NY, 4Department of Radiology, Icahn School of Medicine at Mount Sinai, New York City, NY Purpose/Objective(s): Previous data from our institution suggests that radiographic evidence of extracapsular extension (rECE) in oropharyngeal cancer (OPC) patients predicts for worse distant control and survival. As patients with human papillomavirus (HPV)-related OPC have excellent prognosis, we investigate the impact of rECE on their disease outcomes. Materials/Methods: From our departmental database, 151 cN+ OPC patients with accessible pretreatment CT scans completed definitive or adjuvant radiation therapy ( induction and/or concurrent chemotherapy) from 2005-2013. All CT scans were reviewed by one head/neck radiologist for evidence of rECE. Testing for HPV was performed via p16 IHC and/or HPV DNA PCR for 104 patients (86 HPV+, 18 HPV-). Univariate and multivariate analyses (MVA) were performed to evaluate impact of these factors on locoregional control (LRC), distant control (DC), progression free survival (PFS), cancer specific survival (CSS), and overall survival (OS). Covariates included age, gender, race, smoking history, ECOG performance status (PS), cT-stage, cN-stage, and treatment type. Results: With median follow-up of 28 months (range: 3-93), HPV+ patients had similar rates of rECE as HPV- and untested patients (56% vs 50% vs 55%, p Z 0.902). Patients with rECE underwent definitive chemoradiation (DCRT) more frequently than those without rECE (80% vs 65%, p Z 0.046). HPV+ and HPV- patients underwent similar rates of DCRT (66% vs 61%, p Z 0.786). HPV+ patients were more likely to have lighter smoking history (59% vs 28% 10 pack-years, p Z 0.019), smaller primary tumors (87% vs 39% cT1-2, p<0.001), and trended towards larger nodal burden (86% vs 67% cN2-3, p Z 0.061) vs HPVpatients. For the entire study group, rECE predicted for worse disease control and survival. For HPV+ patients, however, rECE did not significantly impact their outcomes. (Table) On MVA of patients tested for HPV, rECE was found to independently predict for worse LRC (HR 6.70, 95% CI 1.42-31.62) and DC (HR 9.16, 95%CI 1.17-71.78). HPV positivity was found to independently predict for better LRC (HR 0.05, 95%CI 0.010.20) and PFS (HR 0.21, 95%CI 0.09-0.51). Conclusions: HPV+ OPC patients who undergo radiation therapy have an excellent prognosis. Although rECE predicted for worse LRC, DC, and

survival for the entire study group, rECE did not significantly impact outcomes for HPV+ patients. Larger studies with longer follow-up are needed to confirm these findings. J.T. Liu: None. B.H. Kann: None. B. De: None. M. Buckstein: None. R.L. Bakst: None. E.M. Genden: None. M.R. Posner: None. P.M. Som: None. V. Gupta: None.

2768 Temporal Stability and Reproducibility of FDG-PETeBased Dose Painting Targets in Head and Neck Cancer J. Rasmussen,1 B.M. Fischer,2 I.R. Vogelius,1 M.C. Aznar,1 F. Jeppe,1 S.M. Bentzen,3 and L. Specht1; 1Department of Oncology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark, 2Rigshospitalet, University of Copenhagen, PET & Cyclotron Unit, Copenhagen, Denmark, Copenhagen, Denmark, 3Division of Biostatistics and Bioinformatics, University of Maryland Greenebaum Cancer Center, and Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, WA Purpose/Objective(s): The FDG avid sub-volume of the gross tumor volume, assessed on PET scans is hypothesized to be a clinically relevant target in painting strategies. Stability of the FDG positive tumor volume (GTV-PET) is important for clinical implementation of dose painting. The purposes of this study are (1) to assess the stability of FDG PET/CT target volumes in a cohort of patients with HNSCC and (2) to investigate if any potential variability would result in a clinically relevant difference in dose painting plans. Materials/Methods: A total of 29 of a planned 30 patients with oropharyngeal HNSCC were included in this prospective study. All patients were scanned prior to treatment with exactly three days interval between the two scans. Patients were scanned on the same PET/CT scanner, immobilized and according to the same protocol. Standardized Uptake Value (SUV) is corrected for bodyweight. Metabolic activity was measured semi-quantitatively by SUVmax, SUVmean, and the iso-avid MTV50% (volume with FDG avidity >50% of SUVmax) and qualitatively by delineation of GTVPET. The latter was performed blinded by a specialist in nuclear medicine with a delay of 3 months between the delineation of the two scans. Images were co-registered systematically for each patient rigidly on bone (corpus of C2-C5 and the mandible) and DICE similarity coefficients were calculated for MTV50% and will be calculated from GTV-PET (awaiting the 3 months delay). A plan with dose escalation to MTV50% for the patient with the lowest DICE was simulated to estimate the difference in the dose plan. Results: Currently 29 patients are included. The Table shows the mean SUV metrics and MTV50% from the two scans along with the mean difference and DICE coefficient. A plan with dose escalation (79.7 Gy) towards MTV50% from day one and day two was optimized. The mean and min dose to MTV50%1 and MTV50%2 were respectively (80.2, 78.8 Gy and 80.0, 73.8 Gy). Likewise we found no difference in mean dose to PTV- MTV50%1 and PTV- MTV50%2 (79.7 Gy and 78.4 Gy), but a relevant difference in D98% was observed (77.7 Gy and 69.8 Gy). By visual inspection, however, it was clear that this difference was due to the systematic rigid co-registration as mentioned above which in this patient resulted in a poor match probably due to swallowing. The low DICE in this case was furthermore enhanced by a rather small MTV50% (2.5 and 2.8 cm3). Conclusions: The stability of FDG PET/CT metrics in HNSCC patients is good with a high DICE coefficient between PET defined sub-target volumes and likely no clinically relevant impact in dose painting plans.

Volume 90  Number 1S  Supplement 2014 Scientific Abstract 2768; Table

Poster Viewing Abstracts S515

Mean SUV metrics, MTV50% and mean difference from the two PET/CT scans and DICE coefficient

SUVmax1

SUvmax2

Mean difference (SUVmax)

19.11

19.97

0.86

SUVmean1

SUVmean2

Mean difference (SUVmean)

MTV50%1

MTV50%2

Mean difference (MTV50%)

12.43

12.88

0.44

4.35

4.82

0.47

Author Disclosure: J. Rasmussen: None. B.M. Fischer: None. I.R. Vogelius: None. M.C. Aznar: None. F. Jeppe: None. S.M. Bentzen: None. L. Specht: None.

2769 Radiobiological Evaluation of IMRT Treatment of Head and Neck Patients: Multi-institutional Study G. Narayanasamy,1 A. Pyakuryal,2 S. Pandit,3 T.T. Sio,4 J. Vincent,1 M.R. Kudrimoti,5 and Y. Li1; 1University of Texas Health Science Center, San Antonio, TX, 2University of Illinois, Chicago, IL, 3BP Koirala Memorial Cancer Hospital, Bharatpur, Nepal, 4Mayo Clinic, Rochester, MN, 5University of Kentucky, Lexington, KY Purpose/Objective(s): To evaluate and correlate the outcome of the IMRT in head and neck cancer using empirical radiobiological models such as Poisson statistics for tumor control probability (TCP) and JT Lyman model for normal tissue complication probability (NTCP). Materials/Methods: IMRT plans were retrospectively reviewed in this study that was approved by IRB in both the institutions. The TCP and NTCPs were calculated from the dose-volume histogram (DVH) statistics using the HART software (Histogram Analysis in Radiation Therapy). Values for the volume parameter (n), slope parameter (m), tumor control dose (TCD Z 63.8 Gy) and tolerance dose (TD50,5 Z 28.4, 47 and 70 Gy for bilateral parotids, esophagus and larynx, respectively) were selected from Luxton et al. Severity of complications was graded as per Common Terminology Criteria for Adverse Events guidelines (CTCAE ver 4.0). A correlation study was also pursued for tumor progression free survival and toxicity leading to xerostomia, dysphagia. Results: Of the 150 IMRT head & neck patients identified, 30 cases (Rx range: 63-70 Gy) that developed xerostomia (N Z 23) or dysphagia (N Z 19) or both (N Z 13) were studied over a mean follow up time of 1.8 years. TCP of tumor was estimated to be 0.770.05; while NTCP index for parotids, esophagus and larynx were 0.400.34, 0.220.14 and 0.070.14, respectively. Using 2-tailed Student’s t-test on the 30 cases, the severity of dysphagia and xerostomia was found to correlate with the estimated values of NTCP index for esophagus (R2 Z 0.77, p-value<0.0001) and bilateral parotids (R2 Z 0.63, p-value<0.0001), respectively. However JT Lyman model provided poor correlation between severity of dysphagia and NTCP of larynx (R2<0.10). In a sub-group study of 18 patients treated at Rx Z 70 Gy, it was found that Grade 2+ xerostomia toxicity was observed in 33% of patients with NTCP of 0.310.10 for parotid, and Grade 3+ dysphagia toxicity was observed in 22% of patients with NTCP of 0.170.06 for esophagus. Conclusions: In this 1.8 year follow up study on 30 head & neck IMRT treatments across 2 institutions, severity of dysphagia and xerostomia was found to have correlation with the NTCP indices of esophagus and bilateral parotids, respectively. This novel methodology of radiobiological outcome-related analysis may be utilized to evaluate severity of potential complications prior to treatment. Author Disclosure: G. Narayanasamy: None. A. Pyakuryal: None. S. Pandit: None. T.T. Sio: None. J. Vincent: None. M.R. Kudrimoti: None. Y. Li: None.

2770 HPV-Positive Oropharyngeal Cancers: Midtreatment PET During Radiation Therapy and Implications for Treatment De-escalation S. Rosenberg, R. Grewal, N. Riaz, P.B. Romesser, A. Pena, S. McBride, H. Schoder, and N. Lee; Memorial Sloan-Kettering Cancer Center, New York, NY

DICE (Range) 0.72 (0.39-0.93)

Purpose/Objective(s): Patients with HPV+ oropharyngeal squamous cell carcinomas (OSCC) have a better prognosis compared to HPV negative tumors. De-escalation strategies for this population are an active area of investigation. Other disease sites, such as lung cancer (e.g. RTOG 11-06), are attempting to use mid-treatment PET/CT (MT-PET/CT) to guide decisions on treatment intensification. The aim of this study was to describe patterns of response from MT-PET/CTs in HPV+ OSCC and elucidate whether these scans may serve as potential biomarker to identify patients for treatment de-escalation. Materials/Methods: Between 9/2011 and 12/2013, patients with HPV+ OSCC had MT-PET/CTs during their course of definitive IMRT at our center. HPV status was determined by p16 IHC. The majority of patients received concurrent cisplatin-based chemotherapy. A single nuclear medicine physician reviewed all scans. The highest standardized uptake value (SUVmax) was measured for the primary tumor and individual neck lymph nodes (LN). The percent difference in the SUVmax between the two scans was calculated for each primary tumor and LN. EORTC criteria were used to assess response: complete response (CR) was complete resolution of FDG activity, partial response (PR) >25% decrease in SUVmax, stable disease (SD) was a -25% to +25% change in SUVmax, progression of disease (POD) was a >25% increase. Response assessment of the primary and LNs using metabolic tumor volume and tumor-lesion glycolysis will be presented at the meeting. Results: 20 primary tumors and 47 LNs from 23 patients were available for analysis. Primary tumors were excluded if FDG values could not be adequately assessed (e.g. inflammation). 16 patients had a base of tongue primary, while 7 had a tonsil primary. 9 patients had a T1 tumor, 11 a T2 tumor, and 3 a T3-T4 tumor. One patient had N1 disease, the remainder had either 2b or 2c nodal disease. MT-PET/CT was performed at a mean of 42.5 days (range 34 to 52) from start of RT. For the primary, 4 patients had a CR, 13 had a PR, and 3 patients had SD. No patient had progression of the primary. Analysis of lymph nodes revealed that 44.7% had a CR, 38.3% had a PR, 8.5% had SD, and 8.5% had POD. MT-PET/CT primary response did not correlate with T stage (p Z 0.76), N stage (p Z 0.87), smoking status (p Z 0.92), or time to MT-PET/CT (p Z 0.66). Results for LNs were similar. Conclusions: Our preliminary results indicate that HPV+ tumors show a wide spectrum of responses on MT-PET/CT. Response did not correlate with traditional clinical staging variables, suggesting these alone may not be sufficient to guide de-escalation decisions. Further study is warranted to determine if MT-PET/CT can be used to individualize treatment de-escalation decisions based on assessment of a tumor’s responsiveness to therapy. Author Disclosure: S. Rosenberg: None. R. Grewal: None. N. Riaz: None. P.B. Romesser: None. A. Pena: None. S. McBride: None. H. Schoder: None. N. Lee: None.

2771 Intensity Modulated Radiation Therapy (IMRT) Reduces Late Toxicity in Patients With Human PapillomaviruseAssociated (HPV+) Oropharyngeal Carcinoma Treated With Chemoradiation Therapy (CRT) S. Koyfman,1 T.B. Bledsoe,2 J. Barnett,1 C.A. Reddy,1 D. Chute,1 T. Nwizu,1 J.P. Saxton,1 B.B. Burkey,1 J.F. Greskovich,1 and D.J. Adelstein3; 1Cleveland Clinic, Cleveland, OH, 2Yale University School of Medicine, New Haven, CT, 3 Cleveland Clinic Foundation, Cleveland, OH Purpose/Objective(s): Evidence about the benefit of IMRT to reduce serious late effects specifically in pts with HPV+ oropharyngeal squamous cell carcinoma (OPSCC) treated with CRT is sparse. We investigated the impact of IMRT vs standard 3-field radiation therapy (3D-RT) on late effects outcomes in this pt cohort.