Abstracts
S147
1004 POSTER Survival of lung cancer patients residing in the Russian Atomic Shipbuilding Center, Northwest Russia
1006 POSTER The impact of race and socio-economic status on the treatment of early breast cancer in the United States
E. Solovyova1 , M. Nechaeva1 , Y. Skripchak2 , M. Valkov2 . 1 Regional Oncology Hospital, Clinical Oncology, Arkhangelsk, Russian Federation; 2 Northern State Medical Unoversity, Radiology, Radiotherapy and Clinical Oncology, Arkhangelsk, Russian Federation
J. Sariego1 . 1 Aria Health System, Surgery, Philadelphia, PA, USA
Background: Lung cancer is the most common cancer with 1.4 million new cases per year worldwide. Association between ionizing radiation and both incidence of and survival from cancers is well-established. The aim of the study was to assess survival of lung cancer patients in Severodvinsk − the Russian Atomic Shipbuilding Center compared to the regional center of Arkhangelsk and other parts of the region. Methods: Data on all cases of lung cancer (C34) in 2000–2012 were obtained from the Arkhangelsk Regional Cancer Registry. One-and fiveyear survival with 95% confidence intervals (CI) was calculated using life tables method. Independent associations between the place of residence was studied using Cox regression with adjustment for age at diagnosis, stage, gender, and histology. Crude and adjusted hazard ratios (HR) were calculated with 95% CI. Results: Altogether, 1053, 2266 and 4387 cases of LC were registered in Severodvinsk, Arkhangelsk and other parts of the region during the study period. One year survival was 0.40 (95% CI: 0.36–0.42), 0.37 (95% CI: 0.34–0.39) and 0.34 (95% CI: 0.33–0.36), while five-year survival was 0.18 (95% CI: 0.15–0.21), 0.13 (95% CI: 0.12–0.15) and 0.12 (95% CI: 0.11–0.13) in Severodvinsk, Arkhangelsk and other parts of the region, respectively. In crude analysis, patients from Arkhangelsk (HR = 1.10, 95% CI: 1.01–1.20) and other parts of the region (HR = 1.19, 95% CI: 1.10–1.29) had shorter survival than patients from Severodvinsk. After adjustment, the corresponding values were 1.16 (95% CI: 1.06–1.27) and 1.13 (95% CI: 1.05–1.23). Conclusions: The observed differences in survival of LC patents cannot be explained by the differences between settings in age and gender structure. Neither stage at diagnosis was among the explanatory factors. Better socioeconomic status of residents of Severodvinsk compared to other parts of the region including the regional center seems to be the most plausible explanation of better survival of LC patients from Severodvinsk No conflict of interest. 1005 Cancer of unknown primary is associated with diabetes
POSTER
Background: Breast cancer remains one of the most common and frequently treated malignancies in the United States. There has been a body of literature that suggests that treatment across the country varies based upon race and/or socioeconomic status, among other factors. The current analysis was designed to explore the impact of these factors upon the initial treatment offered to patients with early breast cancer. Methods: A retrospective review was undertaken of all early breast cancer cases (Stages 0, I, and II) registered in the American College of Surgeons National Cancer Database Benchmark Reports between the years 2000 and 2011, inclusive. Data were stratified according to race and socioeconomic status. Race was designated either Caucasian or non-Caucasian, with the patients self-identifying. Socioeconomic status was approximated by recording total household income in the following increments: less than $28,000 a year (Group A); from $28000 to $49000 a year (Group B); greater than $49000 a year (Group C). These levels were chosen to coincide loosely with U.S. Census Bureau data; poverty level for an average family of four was estimated as a household family income of less than $28000, while household incomes above $48000 were not considered within poverty guidelines, regardless of the number of dependents. Treatment rendered was either Mastectomy (M) or Breast Conservation Surgery (BCS). Results: A total of 537059 patients with racial data provided were included in the analysis. Of these, 75.1% self-identified as Caucasian and 24.9% as non-Caucasian. There was no statistically significant difference between these groups with regard to the rates of BCS initially offered: 64.7% for Caucasians and 64.1% for non-Caucasians. There was, however, a difference with regard to socioeconomic status (total in this cohort was 511129 with complete income data); 62.6% of Group A patients were offered BCS, versus 63.7% of Group B and 65.9% of Group C patients. This relationship between household income and BCS rates was linear, direct, and statistically significant. Conclusions: Socioeconomic status − as approximated by household income − has a direct influence upon the rates of BCS among early breast cancer patients. At least in the present study, that relationship appears to be financially- rather than racially-based. No conflict of interest.
¨ Center for Primary Health Care X. Li1 . 1 University Hospital of Malmo, ¨ Sweden Research, Malmo,
1007 POSTER Influence of facility size on the treatment of early breast cancer in the United States
Background: Both type 1 diabetes (T1D) and type 2 diabetes (T2D) increase in incidence worldwide. T2D is associated with many cancers. However, no data are available on cancer of unknown primary (CUP), a relatively common, fatal cancer for which tobacco smoking is the only known risk factor. At diagnosis of CUP metastases are found in various organs which has implications for prognosis. We carried out a nationwide study on the association of CUP with T1D and T2D. Patients and Methods: 32,600 T1D patients and 178,000 T2D patients were identified from the national Hospital Discharge Register, Outpatient Register and Primary Health Care Register and these were linked to the Swedish Cancer Registry. Standardized incidence ratios (SIRs) were calculated for CUP from 1997 through 2010 using anyone without diabetes as a reference. Results: The SIR of CUP in 421 diabetic patients was 1.71, highest for CUP with liver (2.17) and respiratory system metastases (1.95). The SIR was 2.91 for T1D but with small number of patients, 1.38 for T2D with insulin treatment and 1.78 for the main group of T2D. CUP with liver and respiratory system metastases were increase for each diabetic type but for T2D also CUP with gastrointestinal and bone metastases were increased. The highest SIRs for CUP were recorded in the three types of diabetes during the first year of follow-up, most likely due to diagnostic bias during medical work-up. Conclusions: The results provide the first demonstration that CUP is one of the cancers associated with diabetes, with definite evidence on T2D. CUP has a poor prognosis, which may be even worse when diabetes is the underlying co-morbidity. A mechanistic question for future work is to resolve whether diabetes promotes primaries that escape detection or their metastatic spread. No conflict of interest.
J. Sariego1 . 1 Aria Health System, Surgery, Philadelphia, USA Background: There has been an increasing trend in the United States in recent years for breast cancer to be treated at the community and local level. As patients have become more informed and more aware, they have demanded quality care at their local hospitals that approaches the care received at larger focused cancer centers. The current analysis was designed to investigate the influence of facility size on the initial treatment rendered to patients with early breast cancer. Methods: A retrospective review was undertaken of all early breast cancer cases (Stages 0, I, and II) registered in the American College of Surgeons National Cancer Database Benchmark Reports between the years 2000 and 2011, inclusive. Facilities were stratified based on number of cases treated annually and/or academic affiliation: Community Cancer Centers (CCC) − 100 to 649 breast cancer cases treated annually; Comprehensive Community Cancer Centers (CCCC) − 650 cases or more treated annually; Teaching/Research Hospitals (T/R) − associated with a university medical school and/or designated as a National Cancer Institute Comprehensive Cancer Care Program. Data were further stratified based on initial treatment offered: Mastectomy (M); or, Breast Conservation Surgery (BCS). Results: A total of 1,641,151 cases were included in the analysis. Of these, 11.9% were treated at CCCs, 59.7% were treated at CCCCs, and 28.4% were treated at T/Rs. Overall, 59.5% of early breast cancer patients underwent mastectomy and 40.5% underwent BCS. At the CCCs, 64.6% of patients were offered breast conservation versus 58.7% at CCCCs and 59.1% at T/Rs. This difference was statistically significant. Conclusions: Contrary to the belief that “newer”, more progressive treatments − such as breast conservation surgery − are rendered at larger teaching facilities, the current analysis demonstrated the opposite. Rather, among the cohort of patients with early breast cancer, BCS was actually offered as initial treatment more frequently at the community level than at larger, academic centers. This is in keeping with the patients’ desire to be
S148 treated if possible at the local level, and it underscores the importance of a viable, progressive breast cancer treatment program at all levels of care. No conflict of interest. 1008 POSTER How to model temporal changes in comorbidity for cancer patients using prospective cohort data M. Van Hemelrijck1 , L. Lindhagen2 , D. Robinson3 , P. Stattin4 , H. Garmo1 . 1 King’s College London, Cancer Epidemiology Group, London, United Kingdom; 2 Uppsala Clinical Research Centre, Statistics, Uppsala, Sweden; 3 Ryhov County Hospital, Urology, Jonkoping, Sweden; 4 Umea University, Surgical and Perioperative Sciences, Urology and Andrology, Umea, Sweden Background: The presence of comorbid conditions is strongly related to survival and affects treatment choice in cancer patients. Increases in comorbidity occur incrementally over time, so traditional time to event analyses are not adequate. Here, we present a method to model temporal changes in Charlson Comorbidity Index (CCI) in cancer patients using data from PCBaSe Sweden, a nation-wide population-based prospective cohort of men diagnosed with prostate cancer. The model is based on the assumption that a change in CCI is an irreversible one-way process, i.e. CCI accumulates over time and cannot decrease. Material and Methods: CCI was calculated based on 17 disease categories with a specific weight (1, 2, 3, and 6) assigned to each category. Disease categories were defined by ICD-codes for discharge diagnoses in the National Patient Register. A state transition model in discrete time steps (i.e. four weeks) was applied to capture all changes in CCI. The transition probabilities were estimated from three modelling steps: 1) Logistic regression model of vital status, 2) Logistic regression model to define any changes in CCI, and 3) Poisson regression model to determine the size of CCI change, with an additional logistic regression model for CCI changes 6. The four models combined yielded parameter estimates to calculate changes in CCI with their confidence intervals. Results: These methods were applied to men with low-risk prostate cancer who received active surveillance (AS), radical prostatectomy (RP), or curative radiotherapy (RT) as primary treatment. There were large differences in CCI changes according to treatment. There was a rapid increase in CCI for men treated with RP or RT. This is likely explained by the fact that in Sweden all medical conditions of a patient are included in the discharge diagnoses following an in-hospital episode. Therefore, conditions that were previously not recorded will occur for the first time after hospitalisation despite the fact the patient may have had this condition for a long time. Furthermore, there was a more rapid increase in CCI for men treated with RT than AS. For instance, at 10 years after treatment a higher proportion of men age 65 had died or changed CCI after RT than after RP, 7.3% (95% CI: 5.1−9.5). The corresponding value for AS versus RP was 4.1% (95% CI: 2.4−5.9). Conclusions: Our method to model temporal changes in CCI efficiently captures changes in comorbidity over time with a minimal number of regression analyses to perform − which would be impossible with tradition time to event analyses. However, our approach involves a simulation step that is not yet included in standard statistical software packages. In our prostate cancer example we showed that there are large differences in development of comorbidities among men receiving different treatments for prostate cancer. No conflict of interest. 1009 POSTER Serum lactate dehydrogenase and survival following cancer diagnosis W. Wulaningsih1 , L. Holmberg2 , H. Garmo2 , H. Malmstrom3 , M. Lambe4 , N. Hammar3 , G. Walldius5 , I. Jungner6 , T. Ng2 , M. Van Hemelrijck2 . 1 King’s College London, Cancer Epidemiology, London, United Kingdom; 2 King’s College London, Division of Cancer Studies, London, United Kingdom; 3 Karolinska Institutet, Department of Epidemiology, Stockholm, Sweden; 4 Karolinska Institutet, Department of Medical Epidemiology and Biostatistics, Stockholm, Sweden; 5 Karolinska Institutet, Department of Cardiovascular Epidemiology, Stockholm, Sweden; 6 Karolinska Institutet, Department of Medicine, Stockholm, Sweden Background: There is evidence that high levels of serum lactate dehydrogenase (LDH) is associated to poorer overall survival in several malignancies, but its link to cancer-specific survival is unclear. Methods: From a large Swedish cohort, a total of 7,895 individuals diagnosed with cancer between 1986 and 1999 were selected for this study. Multivariable Cox proportional hazards regression was used to assess the
Abstracts standardised score and clinical categories of serum LDH prospectively collected within 3 years prior to diagnosis, and the last 3 months preceding diagnosis in relation to overall and cancer-specific death. Site-stratified analysis was performed for major cancers. Analysis for overall cancer was repeated by different lag times between LDH measurements and diagnosis to assess any temporal association with survival. Results: At the end of follow up, 5,799 participants were deceased. Hazard ratios (HR) and 95% confidence intervals (CI) for overall and cancerspecific death in the multivariable model were 1.43 (1.31 to 1.56) and 1.46 (1.32 to 1.61), respectively, for high compared to low LDH measured within 3 years before diagnosis. A stronger association was observed for LDH measured the 3 months preceding diagnosis. Site-specific analysis showed high LDH to correlate with an increased risk of death from prostate, pulmonary, colorectal, gastroesophageal, gynaecological and haematological cancers. Analysis by six-month lag periods showed LDH assessed closer to diagnosis to be more strongly associated with overall and cancer-specific death. Conclusion: Our findings demonstrated an inverse association of baseline serum LDH with cancer-specific survival, corroborating a role of LDH in cancer progression. No conflict of interest. 1010 POSTER Colorectal cancer in young patients: is it a distinct clinical entity? H. Goldvaser1 , O. Purim2 , Y. Kundel2 , O. Ulitsky2 , D. Shepshelovich3 , N. Wasserberg4 , N. Hananel5 , L. Shemesh-Bar6 , T. Shochat7 , A. Sulkes2 , B. Brenner2 . 1 Institute of Oncology, Davidoff Center, Rabin Medical Center, Oncology, peta tikva, Israel; 2 Institute of Oncology, Davidoff Center, Rabin Medical Center, Sackler School of, Oncology, peta tikva, Israel; 3 Rabin Medical Center, Sackler School of Medicine, Medicine A, peta tikva, Israel; 4 Rabin Medical Center, Sackler School of Medicine, Surgery B, peta tikva, Israel; 5 Rabin Medical Center, Sackler School of Medicine, Surgery A A, peta tikva, Israel; 6 Maccabi Health Services, Primary Care Medicine, Rishon Lezion, Israel; 7 Rabin Medical Center, Statistical Consulting Unit, peta tikva, Israel Background: The incidence of colorectal cancer (CRC) in young patients (pts) has increased significantly in the last decade. Several studies, including from our group, have evaluated whether CRC in young pts has unique features, with conflicting results. The aim of our study was to compare various characteristics of CRC between young and older pts, while emphasizing potential differences through a younger age cutoff and a larger database. Material and Methods: This was a single center retrospective case control study. We searched our institutional database for all CRC pts 40 years old or younger who were diagnosed between 1997 and 2013. A control group consisted of consecutive pts older than 50 years at CRC diagnosis during the same period. These pts were matched to the study group by their year of diagnosis in a 1:2 ratio. Pts 41−50 years of age were not included in the study, to accentuate potential age-related differences. Data on demographics, predisposing risk factors, clinical presentation, disease characteristics, treatment and outcome were compared between groups. Results: A total of 330 pts were included: 110 young pts and 220 pts in the control group. The median follow up was 65.9 months (range: 4.7–211). Pts in the younger group had a different ethnic composition, with fewer Ashkenazi Jews (p = 0.0006) and more Arabs (p < 0.0001), increased prevalence of family history of cancer (p = 0.004), hereditary CRC syndromes (p < 0.0001) and inflammatory bowel disease (p = 0.007), and lower incidence of polyps (p < 0.0001).The younger group also had different distribution of tumor location and histological sub-types. They were more likely to present with stage III or IV disease (p = 0.001) and had a higher rate of venous and lymphatic invasion in their tumor (p = 0.012 and p = 0.0006, respectively). They more frequently received treatment, and treatment was more aggressive. The estimated 5-year overall survival (OS) rate were 59.1% and 62.1% in the younger and the control group, respectively (p = 0.57). Interestingly, while the estimated 5-year disease free survival (DFS) was significantly worse for younger pts (57.6% vs. 70%, p = 0.039), young pts with metastatic disease had a strong trend for a better 5-year OS rate (36.5% vs. 16.5%, p = 0.198). Conclusions: Our data suggest that CRC in pts 40 years old or younger at diagnosis may represent a distinct clinical entity. Further research to validate our findings and to define the optimal approach for young CRC pts is clearly needed. No conflict of interest.