Veterans Affairs Central Cancer Registry Research Articles

79 Outcomes of Hepatocellular Carcinoma in the VA Central Cancer Registry

79 Outcomes of Hepatocellular Carcinoma in the VA Central Cancer Registry

Investigating the Risk of Cancer in 1990–1991 US Gulf War Veterans With the Use of State Cancer Registry Data

The purpose of this study was to determine whether proportional cancer incidence is greater among Gulf War veterans compared with non-Gulf War veterans. Files obtained from the Defense Manpower Data Center included data for 621,902 veterans who were deployed to the Persian Gulf during the 1990 to 1991 Gulf War (August 2, 1990, to March 1, 1991) and 746,248 non-Gulf War veteran controls. Identification of veterans who received a cancer diagnosis between 1991 and 2006 was accomplished through record linkage of the Defense Manpower Data Center dataset with files from 28 state cancer registries and the Department of Veterans Affairs Central Cancer Registry. By the use of logistic regression, proportional incidence ratios adjusted for demographic and military characteristics were calculated by comparing the proportion of a specific cancer among all cancers in the Gulf War veterans to the proportion of that specific cancer among all cancers in the non-Gulf War veterans. Only lung cancer showed a statistically significant relative excess among Gulf War veterans compared with non-Gulf War veterans (adjusted proportional incidence ratios, 1.15; 95% confidence interval, 1.03-1.29). When adjusted for race, age, and sex, the overall proportion of cancers among Gulf War and non-Gulf War veterans was similar (odds ratio, 0.99; 95% CI, 0.96-1.02). With the exception of lung cancer, there is little evidence of excess risk of cancer associated with Gulf War deployment. A follow-up study is warranted to confirm this finding and to evaluate the role of greater smoking rates among deployed personnel.

Outcome of Esophageal Carcinoma in the Veteran Affairs Population

Reports indicate that esophageal carcinoma is changing with a shift from squamous cell carcinoma (SCC) to adenocarcinoma (AC) along with worse survival in black patients. However, data on esophageal carcinoma in the Veteran Affairs (VA) population is lacking. We set to describe site-specific and histology-specific characteristics of esophageal cancer in the VA population over the past decades. We queried the VA Central Cancer Registry database for esophageal cancer cases between 1995 and 2005. Patient characteristics and treatment methods were obtained and overall survival was estimated using the Kaplan-Meier method. There were 7929 patients included in the study. The median survival in the period 1995 to 1999 was 7.53 months and 2000 to 2005 was 8.33 months. The median survival in AC was 8.93 months; SCC was 7.30 months. Cox-regression multivariate model revealed that location, age, histology, stage, and treatment were independent predictors of survival. Blacks were less likely to undergo surgery and more likely to undergo radiation. In nonmetastatic esophageal cancer at the VA, SCC has worse outcome as compared with AC. Blacks with AC have worse outcome than whites. Despite equal access to health care blacks are less likely to undergo surgery, which may contribute to their worse outcome. We recommend further studies to determine factors leading to selection of treatment modalities in minority patients.

Outcome of patients with myelodysplastic syndromes in the Veterans Administration population

Purpose Epidemiology and outcome of myelodysplastic syndromes (MDS) in the United States is not well recognized. MDS became reportable to the Surveillance, Epidemiology, and End Results Program (SEER) in 2001. We report first study of MDS among large population in the Veteran Affair system. Patients and methods There are approximately 127 VA Medical Centers diagnosing and/or treating Cancer patients. The data collected by the medical centers cancer registries is aggregated as the VA Central Cancer Registry (VACCR). We used the VACCR to analyze VA patients with MDS diagnosed between 1995 and 2006. The cases were identified using ICD-03 histology codes for MDS. Results A total of 2242 MDS cases were registered during the period analyzed. The median overall survival (OS) was 2.1 years, but varied by French-American-British category. Median OS for patients with RA, RARS and RAEB was 3.4, 4.9, and 0.7 years, respectively. No differences in OS were observed by race. Conclusion Outcome of MDS in the VA was similar to what is described in literature and reported by the SEER Program. Appropriate coding for WHO subtypes, IPSS and treatment details are needed in all MDS registries to facilitate comparisons across populations.

Characteristics of proximal versus distal gastric cancer at the VA: Is there a difference?

15517 Background: Current literature suggests that proximal gastric cancers differ from distal in terms of biological behavior and prognosis. The aim of this study was to assess the differences in the clinical characteristics, treatment and survival in patients with cancers of the proximal versus distal stomach at the VA. Methods: We searched the Veterans Affairs Central Cancer Registry (VACCR) database for stomach cancer between 1995- 2006 using ICD codes 160–169, classified into proximal (cardia and fundus) and distal. The data then was analyzed using SPSS v.13.0. Results: There were 4,251 (68.5%) and 2,045 (32.5%) patients with cancers of the distal and proximal stomach respectively. Baseline characteristics are summarized in table 1. Both Black and white patients were more likely to have cancer of the distal stomach. (p<0.001) The distribution of cancers of the distal and proximal stomach did not change when the data was compared between 1995–2000 and 2001–2006. Metastatic disease was present in 27.3% of proximal vs. 24.9% in distal stomach cancer. (p=0.09) Patients with distal gastric cancer were more likely to undergo surgery and proximal stomach cancer patients were more likely to undergo chemo or radiotherapy (p<0.001), but there was no difference in the median survival. Cox-regression analysis using variables age, race, stage, treatment, only age, chemotherapy, surgery and the absence of metastatic disease were predictors of survival in distal stomach cancer, while surgery, chemotherapy and the absence of metastatic disease were predictors of survival in proximal stomach. Conclusion: Patients of both races were more likely to have carcinomas of the distal than proximal and this had remained stable over the past decade. The survival outcome was similar and uniformly poor in patients with cancer of the proximal and distal stomach. Further studies are required to study the racial variations and the poor prognosis of gastric cancer. Characteristic Number of patients (N = 6,296) P Value Proximal N = 2,045 Distal N = 4,251 Age (Years) 67.29 69.09 P=0.85 Race Whites 1738 (38.4%) 2789 (61.6%) P<0.001 Blacks 307 (17.4%) 1462 (82.6%) P<0.001 Stage (n = 3,606, 57.3%) Non metastatic 741 (37.2%) 1249 (62.8%) P=0.09 Metastatic 558 (34.5%) 1058 (64.5%) P=0.09 Treatment (n = 4,059, 64.5%) No therapy 623 1614 Surgery 522 (27.4%) 1381 (72.6%) P<0.001 Chemotherapy 219 (30.9%) 489 (69.1%) P<0.001 Radiotherapy 139 (48.4%) 148 (51.6%) P<0.001 Survival (Months) 6.9 5.77 P=0.06 Non metastatic disease 12.13 14.27 P=0.18 Metastatic disease 3.63 2.53 P=0.059 No significant financial relationships to disclose.

Ampulla of Vater adenocarcinoma in the Veteran's Affairs population

15521 Background: Outcome data on Ampulla of Vater (AV) cancer is lacking. Surgical resection is the only option for long term survival. We performed a retrospective study of patients with AV adenocarcinoma treated at Veterans Affairs (VA) hospitals. Methods: The VA Central Cancer Registry (VACCR) is a function of the Oncology Program Office at VA Headquarters, Washington DC. We searched VACCR for all diagnosed AV cancer cases (primary site code 241) from 1995 to 2005. Data was entered and analyzed using SPSS v.13.0. Patient and disease characteristics were compared by chi-square test. Overall survival was estimated with Kaplan-Meier method. Multivariate analysis was performed with Cox proportional hazards regression model to determine the independent association between each variable and survival. Results: A total of 367 cases of AV adenocarcinoma were identified. Overall 314 (85.6%) individuals were white, 43 (11.7%) black, and 10 (2.7%) other. Thirty-eight patients (10.4%) had metastatic disease while 180 (49%) did not. Surgery was performed in 210 (57.2%). Overall median survival was 12 months. Survival with metastatic disease was 6.57 months and 14.37 months without metastasis (p=0.0005). In nonmetastatic disease, surgery improved survival from 4.73 to 28.9 months (p<0.001). Patients who received surgery in addition to radiation and chemotherapy for nonmetastatic disease had similar survival of 27.9 months (p=NS). In multivariate analysis with race, metastatic disease, chemotherapy, radiation, and surgery, only surgery was a significant predictor of survival (table 1). Conclusion: Adenocarcinoma of the AV is a rare malignancy of poor prognosis. To our knowledge, a systematic review of a large patient population has not been reported. In the VA, surgery was shown to improve survival in nonmetastatic disease. Adjuvant chemotherapy plus radiation did not add survival advantage. Future studies are essential to determine optimal therapy. Survival in Nonmetastatic Disease Therapy Cases Months None 62 4.73 (1.77–7.7) Surgery 74 28.9 (14.2–43.6) Surgery/Chemo/Rad 22 27.93 (1.75–54.12) No significant financial relationships to disclose.

Intrahepatic versus extrahepatic cholangiocarcinoma: Is there a difference in survival?

15564 Background: Cholangiocarcinoma remains as a therapeutic challenge with surgery often being the only curative option. Recently, it has been described that patients with resectable tumors have median survival of 59 months versus 9 months in patients with unresectable disease. In this retrospective study, we analyze data comparing survival in intrahepatic cholangiocarcinoma (IHCC) versus extrahepatic cholangiocarcinoma (EHCC) across the Veterans Affairs (VA) Medical centers. Methods: The VA Central Cancer Registry (VACCR) is a function of the Chief, Program Office for Oncology at VA Headquarters in Washington DC. We queried the VACCR database for all diagnosed bile duct cancer cases between 1995 and 2005 using primary site code 221 (IHCC) and 240 (EHCC). The data was transformed, entered and analyzed using SPSS v.13.0. Patient and disease characteristics were compared using chi-square test. Overall survival was estimated with Kaplan-Meier method. Results: A total of 825 cases of cholangiocarcinoma were identified, of which 340 (41.2%) were IHCC and 485 (58.8%) were EHCC. Overall 122 (14.8%) patients were black (IHCC=52, 15.3%; EHCC=70, 14.4%; p=0.123). In those with IHCC, 78 (22.9%) had metastatic disease vs. 80 (16.5%) in EHCC (p<0.001). Surgery was more likely to be performed in EHCC than IHCC (26% vs. 9.1%; p<0.001). Median survival in EHCC was 5.97 months vs. 3.67 months in IHCC (p<0.001). Surgery significantly increased overall survival in both groups (table). Conclusion: IHCC and EHCC remain as therapeutic challenges. Surgical resection is the mainstay of therapy in VA patients with cholangiocarcinoma and has been shown to increase survival of these patients. Although surgical resection was performed more often in EHCC than IHCC, survival was similar in both if surgical resection was achieved. Surgery did increase overall survival in both groups. Overall Survival in IHCC and EHCC Survival (months) IHCC EHCC p-value All cases 3.67 5.97 <0.001 Metastasis 2.17 2.93 0.45 No Mets 5 6.57 0.374 Surgery 13.63 13.47 <0.001 No Surgery 3.13 4.4 <0.001 No significant financial relationships to disclose.

Myelodysplastic Syndrome: A Study of VA Population.

Background: Epidemiology and outcome of myelodysplastic syndromes (MDS) in the United States is not well recognized. MDS became reportable to the Surveillance, Epidemiology, and End Results program (SEER) in 2001. Only one study is published examining the SEER data between 2001–2003 on MDS incidence, epidemiology and outcome. We report first study of MDS among large population in the Veteran Affair system.Methods: There are approximately 127 VA Medical Centers diagnosing and/or treating Cancer patients. The data collected by the medical centers cancer registries is aggregated as the VA Central Cancer Registry (VACCR) and is maintained by the Chief Program Office for Oncology at VA in Washington DC. We used the VACCR to analyze VA patients with MDS diagnosed between 1995 and 2005. The data was downloaded from the database using ICD-03 histology codes for MDS. Data was entered and analyzed using bio-statistical software SPSS. Kaplan-Meier curves were used for estimates of overall survival, chi square for comparison of categorical variables and t-test for continuous variables.Results: Between 1995 and 2005, 1411 MDS cases were registered in the VACCR database. The median age was 75 years. Among those 1379 (97.7%) were males and 1196 (84.4%) were white patients, 171 (12.1%) black, and 44 (3.5%) other race. Majority of cases were reported as MDS, NOS 935 (66.3%), Refractory anemia (RA) 180 (12.8%), Refractory anemia with ring sideroblasts (RARS) 105 (7.4%), Refractory anemia with multilineage dysplasia (RCMD) 22 (1.6%), 5 q syndrome 13 (0.9%), Refractory anemia with excess blasts (RAEB) 105 (7.4%), Refractory anemia with excess blast in transformation 17 (1.2%), and therapy related MDS (t-MDS) 34 (2.4%). No IPSS data were available. The median overall survival (OS) was 1.8 year (95% CI 1.5–2.1). The 3 year and 5 year overall survival was 30% and 22% respectively. The median OS for RA patients was 3.7 year, RARS 4.9 year and for RAEB was 0.8 year. No difference was observed in OS between whites and black (median OS 1.73 versus 2.65, p value 0.28). Their was a non statistically significant trend for improvement of OS in patients diagnosed in 2004–2005 compared to 1995–2003 (time where azacitidine and lenalidomide were approved) median OS of 3.2 versus 1.8 (p value 0.64).Conclusion: This is the first report of MDS case series in the VA system. Outcome is similar to what is described in literature and SEER data (3 Year OS was 30% in our VA population study compared to 35% in recently published SEER data). No racial disparities are observed in outcome. A trend towards better overall survival is noted in the last 2 years since approval of new treatment options. Coding for WHO subtypes, IPSS and treatment options should be considered in all large MDS registries.

Improvement in Follicular Non Hodgkin Lymphoma Survival among VA Patients.

Background: Follicular non Hodgkin lymphoma (FL) remains uncurable disease for majority of patients. Natural history of follicular lymphoma is changing with inroduction of new treatments. We examined outcome of FL among VA patients in the era of monoclonal antibodies.Methods: This was a retrospective analysis. The VA Central Cancer Registry (VACCR) database was used to identify patients with FL diagnosed between 1995 and 2005. There are approximately 120 VA medical centers diagnosing and/or treating patients with cancer. Data were analyzed using bio-statistical software SPSS. Variables included age, sex, stage of disease, histology subtype, date of diagnosis, date of last contact, vital status.Results: There were1338 patients with FL at the VACCR database between 1995 and 2005, 283 (88%) were white, 128 (10%) black patients, and 39 (2%) patients from other racial groups. The mean age at diagnosis was to 64 years. Four hundred twelve (31%) were grade I FL, 312 (23%) Grade II, 215 (16%) Grade III, and 399 (30%) FL, NOS. Four hundred twelve (30%) patients were early stage (I,II), 600 (45%) were advanced stage (III,IV) and 331 (25%) missing or unkown. No FLIPI score data were available. Median overall survival (OS) was 7 years. The median OS for early stage was 8.6 years compared to 5.5 years for advanced stage. The median OS for patients diagnosed 1995–2000 was 5.9 years compared to 9.7 years for patients diagnosed between 2001–2005. (log rank test, P value 0.025). The two groups were balanced regarding stage, race and chemotherapy.Conclusions: The survival of FL patients has improved in the VA system in the era of rituximab. [Display omitted]

The relationship of lymph node dissection and colon cancer survival in the Veterans Affairs Central Cancer Registry

Background The extent of lymphadenectomy in colon cancer may impact potential to cure and accuracy of staging. Methods The Veterans Affairs Central Cancer Registry database was queried for TNM stage I–III colon adenocarcinoma patients and yielded 5,823 individuals. The number of lymph nodes examined, number positive, and the positive:examined lymph node ratio were studied with respect to overall survival by using log-rank and Kaplan-Meier analysis. Results The overall survival (OS) in stage II patients was greater with the higher number of lymph node (LN) examined. For stage II patients, the 5-year OS was 34%, 43%, 47%, and 55% for the lowest to highest quartiles ( P = .007). For stage III patients, the 5-year OS was 31%, 27%, 38%, and 53% for the lowest to highest quartiles (not significant overall). OS is greater with an increased number of positive lymph nodes ( P < .001). The lymph node ratio was more powerful prognostically with a 5-year OS of 27% for the highest quartile versus 44% for the lowest. Conclusions More extensive lymphadenectomy is associated with improved OS in stage II colon cancer patients. The positive:examined LN ratio is more powerful prognostically than the number of nodes examined or LN positivity.

Esophageal squamous cell carcinoma in the VA population: What is the optimal treatment modality

15042 Introduction: Surgery for esophageal squamous cell carcinoma (SCC) has been the mainstay of treatment despite dismal outcome and significant surgical complications. There is no standard treatment modality for esophageal SCC. Methods: The VA (veteran affairs) Central Cancer Registry (VACCR) is a function of the Chief, Program Office for Oncology at VA Headquarters in Washington DC. We queried the VACCR database for all diagnosed squamous cell esophageal cancer cases between 1995 and 2005 using ICD codes 150–159. The data was transformed, entered and analyzed using SPSS v.13.0. We analyzed, in a retrospective fashion, survival in VA patients with early disease (stages 1–2), and locally advanced (stage 3) SCC comparing the treatment modality: chemoradiation alone (CRT), surgical resection alone (SUR) or trimodality therapy (TMT) which includes all three treatment options. Results: Out of a total of 6874 patients diagnosed with esophageal carcinoma, 2894 patients had SCC. A total of 433 patients were included in this study that were staged as 1–3 and had complete treatment information available. Baseline characteristics were not different between the three groups and are summarized in table 1 . Out of those, 57 (13.2%) received SUR, 323 (74.6%) CRT, and 53 (12.2%) TMT. Kaplan Meier analysis for median survival in early disease was 14 mo for SUR, 17 mo in CR, and 79 mo in TMT (p = 0.0288). There was no difference in survival among patients with locally advanced disease between the treatment groups (p = 0.7079) Conclusion: In VA patients with early esophageal SCC, TMT confers better survival than SUR or CRT. However, in patients with advanced disease, SUR, CRT and TMT groups showed comparable outcome. [Table: see text] No significant financial relationships to disclose.

Esophageal cancer at the VA: Does histology matter?

15069 Introduction: In recent years esophageal cancer has shown a shift from largely squamous cell carcinoma (SCC) histology to adenocarcinoma (AC). This is felt to be due to changing risk factor profiles among the American population. Methods: The VA Central Cancer Registry is a function of the Chief, Program Office for Oncology at VA Headquarters in Washington DC. We queried the VACCR database for all diagnosed esophageal cancer cases between 1995 and 2005 using ICD codes 150–159. The data was transformed, entered and analyzed using SPSS v.13.0. Results: There were a total of 6874 cases diagnosed between 1995 and 2005. Out of those, 2968 (43.2%) were diagnosed with AC and 2894 (42.1%) with SCC. Black patients were more likely to have SCC than AC. Baseline characteristics are summarized in table 1 . Staging information was available in 2823 (48.1%) patients. Patients with AC were more likely to present with metastatic disease than those with SCC (52.7% vs. 44.3%). Median survival in AC was better than SCC for non-metastatic disease (12.3 mo vs. 8.6 mo, p=0.0009). No difference in survival was seen in metastatic disease between the two histologies (AC 4.2 mo vs. SCC 3.97 mo p=0.537). In Cox- regression analysis using variables including race, histology, stage, grade and treatment, only histology, stage, grade and treatment were significant predictors of survival. Conclusions: In non-metastatic esophageal cancer at the VA, SCC has worse outcome compared to AC. However, in metastatic disease, SCC and AC have similar outcome. We recommend further studies to determine if the differences in histology are determined by biological or lifestyle differences. [Table: see text] No significant financial relationships to disclose.

Male breast cancer in the veterans affairs population

The incidence of male breast cancer (MBC) continues to rise. The Veterans Affairs (VA) Central Cancer Registry (VACCR) provides a unique source for the study of MBC. The objective of this retrospective analysis was to compare the characteristics and outcome of patients with MBC and patients with female breast cancer (FBC) in the VA population. VACCR data were used to analyze the database of VA patients who had breast cancer diagnosed between 1995 and 2005. It includes 120 VA medical centers. Primary site codes were identified for breast cancer (500-508). Data were entered and analyzed using biostatistical software. In total, 3025 patients' records were reviewed, and 612 patients who had MBC were compared with 2413 patients who had FBC. The mean age at diagnosis was 67 years for patients with MBC and 57 years for patients with FBC (P < .005). More patients with MBC were black, and patients with MBC presented with higher disease stage and more lymph node-positive disease. The dominant histology in MBC was ductal carcinoma. No difference in grade or laterality was observed. Estrogen and progesterone receptor-positive tumors were more common in MBC compared with FBC. Overall, patients with MBC received less chemotherapy, whereas no statistical difference was observed in the use of hormone treatment. The median overall survival for patients who had MBC was 7 years compared with 9.8 years for patients who had FBC (log-rank test; P < .005). There was no statistically significant difference in median survival for patients with stage III disease and stage IV disease. However, the median survival differed significantly for patients with stage I disease and stage II disease. In lymph node-negative patients, the median survival was 6.1 years for patients with MBC and 14.6 years for patients with FBC (P < .005), whereas the median survival did not differ significantly in lymph node-positive patients. Using Cox regression analysis age, sex, clinical stage, and lymph node status were independent prognostic factors for survival, whereas race, histology, and grade were not. To the authors' knowledge, this is the largest series of MBC and FBC to date in the veterans population. The results suggested the presence of differences in the biology, pathology, presentation, ethnicity, and survival between patients with MBC and patients with FBC in the VA population. It is noteworthy that the survival of patients with MBC was inferior for those with early-stage disease and lymph node-negative tumors, suggesting that there are differences between the sexes in the pathogenesis and biology of breast cancer. In patients with hormone receptor-positive MBC, survival was inferior despite similar hormone treatment practices between MBC and FBC. This observational study calls for a better understanding of this disease that would allow new opportunities for specific therapeutic intervention.

P63: The relationship of lymph node dissection and extrahepatic cholangiocarcinoma in the VA central cancer registry

P63: The relationship of lymph node dissection and extrahepatic cholangiocarcinoma in the VA central cancer registry

Outcome of Diffuse Large B Cell Lymphoma in the VA Health System: Do We Need To Do Better?.

Background: Diffuse Large B cell lymphoma (DLBCL) is the commonest subtype of non-Hodgkin Lymphoma (NHL). Data on outcome of DLBCL in large cohorts of patients or community practices outside clinical trials are lacking. Treatment and outcomes of DLBCL among the VA patients have not been reported previously. We utilized the VA Central Cancer Registry to study cases of DLBCL in the VA system.Methods: This was a retrospective analysis. The VA Central Cancer Registry (VACCR) database was used to identify patients with DLBCL diagnosed between 1995 and 2005. There are approximately 120 VA medical centers diagnosing and/or treating patients with cancer. The VACCR aggregates the data collected by the medical centers' cancer registries. Data were extrapolated and analyzed using bio-statistical software SPSS. Variables included age, sex, stage of disease, histology subtype, date of diagnosis, date of last contact, date of relapse, vital status, whether patients received chemotherapy and or radiation.Results: There were 2792 patients with DLBCL at the VACCR between 1995 and 2005, 2402 were white, 323 black patients, and 67 patients from other racial groups. The mean age at diagnosis was to 65. Majority were males (2721 patients). More patients were diagnosed at advanced stage (stage III and IV) (40%), in 34% of cases stage was not known or missing. IPI score data were not available. Most patients received multi-agent chemotherapy 70%, and 20% of patients received radiation (31 % in early stage and 12% in late stage). The 5-year overall survival was 28%. 5 year OS was 41% for stage I, 32% for stage II, 25% for stage III and 17% for stage IV. Using Cox regression multivariable analysis age and use of mutli-agent chemotherapy were the only statistically significant independent variables affecting survival.Conclusions: Outcome of DLBCL in VA patients seems inferior to what is reported in clinical studies. The age of patients treated in the VA as well as sex could be factors affecting the outcome. There are no data available on outcome of DLBCL in such large cohort of patients outside context of clinical trials to make meaningful comparisons.

Outcome of Diffuse Large B Cell Lymphoma in the VA System: The Rituximab Era.

Background: Addition of rituximab to chemotherapy in patients with diffuse large B cell Lymphoma (DLBCL) has been shown to improve survival in several recent clinical studies. A study from British Columbia confirmed those results in a population-based cohort. No similar population based studies were conducted in the USA. Our study aims to address outcome of DLBCL in the era of rituximab in the VA health system.Methods: This was a retrospective analysis. The VA Central Cancer Registry (VACCR) database was used to identify patients with DLBCL diagnosed between 1995 and 2005. There are approximately 120 VA medical centers diagnosing and/or treating patients with cancer. The VACCR aggregates the data collected by the medical centers' cancer registries. Data were extrapolated and analyzed using bio-statistical software SPSS. Variables included age, sex, stage of disease, histology subtype, date of diagnosis, date of last contact, date of relapse, vital status, whether patients received chemotherapy and or radiation. Use of rituximab was not specifically recorded in the registry. Due to that we divided the patients into two groups, patients diagnosed with DLBCL before 2001 (pre rituximab era group) and patients diagnosed after 2001 (rituximab era group). The initial results of rituximab in DLBCL were presented in 2000 and published in January 2002. Independent t test was used for comparing continuous variables and chi square test for categorical variables. Wilcoxon test was used to compare survival among the two groups.Results: There were 2792 patients with DLBCL at the VACCR between 1995 and 2005, 1772 patients in pre rituximab era and 1020 patients in the rituximab era. The mean age at diagnosis was to 64 in pre rituximab group and 66 in rituximab group (P-value 0.015). Race distribution was similar between the two groups. More patients were diagnosed at advanced stage (stage III and IV) 61 % in rituximab group compared to 57% in pre rituximab group (P-value <0.005), IPI score data were not available. More patients in pre rituximab era did not receive multi-agent chemotherapy 28% versus 22% (P-value < 0.005). More patients received radiation 21 % in pre rituximab group compared to 16% in rituximab era group (P-value < 0.005). The 5-year overall survival was 26% in pre rituximab era and 36% after rituximab (P-value 0.0025). Using Cox regression multivariable analysis age, use of mutli-agent chemotherapy, radiation and whether patients were diagnosed and treated before or after 2001 were statistically significant independent variables affecting survival.Conclusions: Overall survival of DLBCL in VA patients had improved in the rituximab era. The magnitude of improvement observed in this study is similar to what was described in previous studies. Other factors contributing to improvement in outcomes such as supportive care could not be differentiated in this study. This is a population-based study suggesting improvement in survival in DLBCL in the rituximab era.

Racial Disparities in Diffuse Large B Cell Lymphoma: A VA Population Study.

Background: Racial disparities in diffuse large B cell non-Hodgkin's lymphoma (DLBCL) are not well studied. Racial differences in outcome are not well recognized given majority of patients with DLBCL are white. The VA health care system offers a platform to study those racial differences given similar other variables and socioeconomic status among those patients.Methods: This was a retrospective analysis. The VA Central Cancer Registry (VACCR) database was used to identify patients with DLBCL diagnosed between 1995 and 2005. There are approximately 120 VA medical centers diagnosing and/or treating patients with cancer. The VACCR aggregates the data collected by the medical centers' cancer registries. Data were extrapolated and analyzed using bio-statistical software SPSS. Variables included age, sex, stage of disease, histology subtype, date of diagnosis, date of last contact, date of relapse, vital status, family history of cancer, tobacco history, alcohol history, Agent orange exposure and whether patients received chemotherapy and or radiation. Independent t test was used for comparing continuous variables and chi square test for categorical variables. Wilcoxon test was used to compare survival among the two groups.Results: There were 2792 patients with DLBCL at the VA system, 2402 white and 323 black patients. Sixty-seven patients from other racial groups were excluded from this analysis. The mean age of presentation among blacks was 57.8 years compared to 65.7 years among whites (P-value < 0.005). More Black patients had history of alcohol use, and smoking. More white patients had family history of cancer. No differences in histology subtypes were observed. Other baseline characteristics were similar. No difference in stage of disease was noted at presentation. IPI score data were not available. Similar proportion of patients received multi-agent chemotherapy among the two groups, 70 % for blacks and 67% for whites (P-value 0.58). Among blacks 16 % of patients received radiation while 19.5 % white patients did (P-value 0.73). The 5-year overall survival for blacks was 24 % compared to 29% for whites (P-value 0.92, Wilcoxon test). No statistically significant differences in survival were noted between the two races within each clinical stage. Using Cox regression multivariable analysis age, stage and chemotherapy were the only statistically significant independent variables affecting survival.Conclusions: DLBCL is less common among blacks. Blacks diagnosed with DLBCL are more likely to present at younger age. The treatment and outcome of DLBCL among blacks are not different from whites within the VA system.

Hodgkin’s lymphoma in the elderly veterans: We need to do better

17554 Background: Recent studies demonstrated worse outcome in elderly patients diagnosed with Hodgkin’s Lymphoma (HL). The purpose of this study was to compare the presenting features, treatment, and outcome of elderly patients with HL within the Veteran Affairs Health care system (VA). Methods: This was a retrospective analysis. The VA Central Cancer Registry (VACCR) database was used to identify patients with HL diagnosed between 1995 and 2005. There are approximately 120 VA medical centers diagnosing and/or treating patients with cancer. Data are entered by tumor registrars at the VA medical centers. This site aggregates the data collected by the medical centers’ cancer registries. Data was extrapolated and analyzed using bio-statistical software SPSS. Results: We analyzed 1009 patients in the Veteran Affairs (VA) database with the diagnosis of Hodgkin’s lymphoma. Patients were divided into two groups according to age, elderly group (&gt; 60 years) (n = 383) and young group (Age ≤ 60 years) (n = 626). The median age of presentation in elderly patients was 70 years; the most common stage at diagnosis was Stage IV. The most common histological subtype was nodular sclerosis and bulky disease was present in 89 (23.2%) of patients. Out of 383 patients 237 (62%) were treated with chemotherapy, 88 patients (23%) received radiation. Compared to patients ≤ 60 years of age, elderly patients were less likely to be current smokers, more likely to be caucasian, all other baseline characteristics were not statistically significantly different. Less elderly patients received chemotherapy compared to young patients (62% versus 74%) (P-value 0.004). No difference was observed in radiation therapy. The 5 year Overall survival in the elderly was 27 % compared to 70% in patients aged 60 or less. (P-value &lt;0.005). Conclusions: Outcome of HL among elderly veterans is poor. Elderly patients received less chemotherapy treatment. Novel strategies to better deliver treatment in elderly patients with HL are warranted. No significant financial relationships to disclose.

Hodgkin’s lymphoma among veterans: Does race matter?

17538 Background: Racial disparities in Hodgkin’s lymphoma (HL) are not well studied. Earlier studies suggested that black patients received less treatment and had worse outcome. The VA health care system offers a platform to study those racial differences given similar other variables and socioeconomic status among those patients. Methods: This was a retrospective analysis. The VA Central Cancer Registry (VACCR) database was used to identify patients with HL diagnosed between 1995 and 2005. There are approximately 120 VA medical centers diagnosing and/or treating patients with cancer. Data are entered by tumor registrars at the VA medical centers. This site aggregates the data collected by the medical centers’ cancer registries. Data was extrapolated and analyzed using bio-statistical software SPSS. Results: There were 1009 patients with HL at the VA system, 801 white and 183 black patients. Twenty-five patients from other racial groups were excluded from this analysis. The median age of presentation among blacks was 50.8 years compared to 53.8 years among whites (P-value 0.026). More white patients had family history of HL. Black patients had a higher incidence lymphocyte depleted histological subtype 33 (18.0%) vs. 75 (9.4%) and whites had a higher prevalence of nodular sclerosing subtype, 260 (32.5%) vs. 32 (17.5%). Other baseline characteristics such as sex, history of smoking, alcohol, and radiation exposure were similar. Presence of bulky disease and stage of lymphoma at presentation were also similar. Similar proportion of patients received chemotherapy among the two groups, 70 % for blacks and 69% for whites. Among blacks 19.7 % of patients received radiation while 25.5 % white patients did (P-value 0.24). The 5-year overall survival of blacks was 53 % years as opposed to 54% for whites (P-value 0.93) Conclusions: The presenting features, risk factors are similar across black and white races among VA population. Blacks are more likely to present with lymphocyte depleted type of HD. The treatment and outcome of HL among blacks are not different from whites within the VA system. No significant financial relationships to disclose.

Do racial differences predict survival amongst VA patients with colon cancer?

3628 Background: Several reports suggested that African-American patients with colon cancer have poor survival. The outcome of colon cancer in patients treated within the VA system is not well studied. Patients treated in the VA usually share common features and socioeconomic class that will allow studying any potential effect of racial difference on outcome. Methods: We used the VA Central Cancer Registry (VACCR) to analyze VA patients with colon cancer diagnosed between 1995–2005. The reference date for data collection and reporting is January 1, 1995. Data are entered by tumor registrars at the VA medical centers. This site aggregates the data collected by the medical centers’ cancer registries. Data was analyzed using bio-statistical software SPSS. Results: There were a total of 14,816 cases with invasive adenocarcinoma.Majority of patients (98%) were males. Median age was 69 years among caucasian patients as compared to 67 years among African- American (p-value &lt;0.005). There was no difference between the two groups with regards to sex, histological grade, histological subtype, use of chemotherapy or radiation therapy. However, African-Americans had less definitive surgery (84% vs. 86%) (p-value 0.003).The staging information was missing or unknown in 59% of the patients. In patients with known stage, no difference was observed.Median survival for caucasians was greater than African-Ameircans (42 months vs. 39 months)(p-value 0.009). Median survival for stage IV patients was not statistically significant between the two groups (8.8 months vs. 8.6 months);(p-value 0.53). Race was not statistically significant independent variable in Cox multiple regression analysis. Conclusions: In the VA system, African-American patients with colon cancer seem to have a trend towards inferior survival. However, race was not an independent prognostic variable. This slight difference could be due to other variables like definitive surgery. No significant financial relationships to disclose.