PURPOSE: Myelodysplastic syndrome (MDS) is a group of clonal bone marrow disorders leading to bone marrow failure and acute myeloid leukemia. With better understanding of the biology of MDS, newer treatment options like therapy with the hypomethylating agents, azacitidine and decitabine have proven beneficial. Major randomized controlled trials by CALGB and others have demonstrated the efficacy of azacitidine. Growth factor use was prohibited in these trials. However, in the community, most patients have continued to get growth factor support along with azacitidine therapy. Since there is lack of evidence backing such clinical practice, we attempted to review response to therapy in patients receiving combination therapy with azacitidine and erythropoietin. METHODS: We identified MDS patients treated with concomitant azacitidine and erythropoietin at our institution over the last five years. Data collected from retrospective chart review of these patients was statistically analyzed. RESULTS: Sixteen patients with concomitant azacitidine and erythropoietin therapy were identified. Complete remissions were seen in 31%(5/16) of the patients. Induction of partial remissions in our study was 25% (4/16). Thirty one percent (5/16) of patients had hematologic improvement not meeting criteria for partial remission. Thirteen percent of patients (2/16) did not benefit at all from azacitidine and erythropoietin support. The median number of cycles received for adequate response was six. A total response (complete remission, partial remission and hematological improvement) of 87%(14/16,) was seen with combination therapy, which is significantly higher than previously published total response of 44–60%. CONCLUSION: Azacitidine with erythropoietin support provides important clinical benefits for patients with MDS. There was a trend towards a higher rate of complete and partial responses in these patients than reported earlier in patients who received azacitidine alone. Randomized controlled studies will be needed to validate this further. Comparison of response to therapy in high risk MDS: Responses *CALGB-9221(azacitidine) **Combined analysis(azacitidine) Study patients treated with azacitidine and growth factors *JCO May 15 2002: 2429–2440.**JCO Aug 20 2006: 3895–3903. Complete Remission 7% 10–17% 31% Partial Remission 16% Rare 25% Hematological Improvement 37% 23–36% 31% Total response 60% 44% 87%

No abstract

e17017 Background: Positron emission tomography (PET) scanning is widely used in staging of patients (pts.) with head and neck squamous cell cancers (HNSCC). However, the role of PET scan in pts. with early stage disease and clinically negative nodes remains controversial. The purpose of our study was to correlate preoperative PET scan findings in early stage disease with operative pathological findings of neck dissection. Methods: A retrospective review of medical records of all pts. treated at Long Island Jewish Medical Center with T1N0 HNSCC from 2002 to 2008 was undertaken. Institutional Review Board approval was obtained for this study. Eligibility criteria included: clinically staged T1N0, documented pre-operative PET scans and primary surgical management including a neck lymph node dissection as part of initial treatment. Data collected included age, gender, physical examination findings, PET scan results, operative pathology and modalities of treatment. Results: Sixty six pts with early stage T1N0 HNSCC were identified. Twelve of sixty six pts met eligibility criteria for this study. Median age was 61 years (range: 27 to 84 yrs), M:F ratio was 1:1. Primary sites of tumor were: tongue (n = 8) 67%; oral cavity (n = 3) 25%; larynx (n = 1) 8%. Preoperative PET scan was positive at primary sites in all pts (n = 12). Additionally, PET positivity was noted in regional lymph node sites in six of twelve pts. Operative pathology confirmed tumor presence in all primary sites. However, nodal involvement was not found in any pt. Median numbers of lymph nodes dissected were 25 (range:13–34). Based on these clinical data, the calculated statistical parameters for preoperative PET scan in detecting occult neck lymph node metastases were: negative predictive value: 100%; positive predictive value: 50%: specificity: 50%; sensitivity: 0%. Conclusions: The clinical application of PET scans in early stage (T1N0) HNSCC may be limited by its high false positive rate. A high negative predictive value suggests that a negative result can exclude metastatic lymph nodes involvement in this group of pts. with early stage disease. Correlation with SUV values of the PET scans and comparisons between the relative PET avidity of the primary tumors and the nodal uptake are ongoing. No significant financial relationships to disclose.

16523 Background: Malignant salivary gland tumors comprise 3 to 6% of Head and Neck cancers. Long term survival data for salivary gland tumors are lacking. We undertook an analysis of all such patients treated at our institution over a 15 yr period. Methods: IRB approval was obtained for this retrospective analysis. Tumor registry data were reviewed for all adult patients diagnosed with salivary gland malignancies for the years 1990 - 2005. Age, Gender, Stage, histology, treatment modalities and survival data were recorded. Results: 200 patients were identified. Median age was 64 yrs (59–69). M:F ratio was 1:1.1 (M: 96; F: 104). Histology and stages: ( Table 1 )Mucoepidermoid Carcinoma (n=50), Adenoid Cystic Carcinoma (n=21), Acinar Cell Carcinoma (n=19), Adenocarcinoma (n=22; Stages: I= 7, II= 4, III= 3, IV= 8), Squamous (n= 12; Stages: I= 1, II= 3, III= 3, IV=5), Poorly Differentiated Carcinoma(n=10), Epithelial Myoepithelial Carcinoma (n=7), Malignant Mixed Carcinoma (n=4), Malignant Myoepithelioma (n=2), Carcinoma in Pleomorphic Adenoma (n=2) and Others (n=15). Hodgkin's and NHL (n=36).Treatment modalities: Surgery (S): n=88 (44%), Radiotherapy (RT): n=6 (3%); Chemotherapy (CT): n=6 (3%); S+RT: n =76(38%); S+CT: n=8 (4%); RT+CT: n =2(1%); S+RT+CT: n =4 (2%). Observation only : n=10(5%). Survival data (Med. survival and 5 yr OS respectively) for all histologies excluding lymphomas are: Stage I - 84 mos,93%; II - 93 mos,85%; III - 39 mos,60 %; IV - 24 mos, 40%. Survival data (Med survival and 5 yr OS) for lymphomas: Stage I - 55 mos,85%; II - 20 mos, 0%; III - 100 mos, 100%; IV- 48 mos, 25%. Median survival of all histology types excluding lymphoma by treatment: S=55 mos, S+RT= 60 mos, S+ CT= 56 mos, RT= 53 mos, RT+CT=15 mos, CT= 24 mos. Conclusions: This large series provides long term outcome data for a relatively rare group of HNCs. Long term survival is noted in several histological sub-types even in the setting of advanced disease. These data should help further define the natural history and biological behaviors of these tumors. [Table: see text] No significant financial relationships to disclose.