Search

Dietary patterns that involve both a decrease in fat and an increase in fruit and vegetable (FV) intake may decrease cancer risks. In this study, a total of 122 premenopausal women with a family history of breast cancer were randomized into one of four diets for 12 mo: nonintervention, low-fat (15% of energy from fat), high-FV(9 servings/d), and combination low-fat/high-FV Fasting blood samples were obtained at baseline and after 3, 6, and 12 mo. Levels of 8-isoprostane-F2a in plasma were deter-mined by immunoassay kits. Statistical analyses indicated that levels of 8-isoprostane-F2a decreased significantly with time in the low-fat arm, which is the only intervention that resulted in weight loss; there were no significant changes in the other three diet arms. It is unlikely that this is due to changes in levels of blood lipids because there was little change overtime in plasma total cholesterol, high-density lipoprotein,low-density lipoprotein (LDL), or triglyceride levels in any diet arm, although mean LDL did decrease slightly in women who reduced fat intake after adjustment for change in body mass index (BMI). Levels of baseline 8-isoprostane-F2a were significantly higher in obese women than in overweight or normal weight women, and change in BMI was significantly correlated with change in 8-isoprostane-F2a levels. These results indicate that low-fat diets or high-FV diets are unlikely to affect plasma levels of 8-isoprostane-F2a in healthy,premenopausal women who do not lose weight during dietary change.

Oxidative DNA damage in blood appears to be useful as a marker of systemic oxidative stress levels. Dietary factors such as fat and energy intakes have been indicated to affect oxidative stress levels, and this may be an important mechanism by which diet can modulate cancer risk. The primary objective of this study was to investigate the effects of dietary intervention in premenopausal women on the levels of one type of oxidative DNA damage: 5-hydroxymethyl-2'-deoxyuridine. The trial randomly assigned women to control, low-fat, low-energy or combination low-fat/low-energy diets for 12 weeks. Blood samples were obtained every 2 weeks, and DNA was analysed for the levels of 5-hydroxymethyl-2'-deoxyuridine. Levels of DNA damage declined with time in each diet arm, including the control arm. The decreases were greater in the two arms with low-energy intake, but not significantly so. The numbers of women who exhibited decreased 5-hydroxymethyl-2'-deoxyuridine levels at 12 weeks versus baseline levels, however, was significantly greater in women assigned to any intervention diet (79%) than in the control arm (50%). Low-fat and low-energy diets therefore had a small effect on changes in oxidative DNA damage levels. The women participating in this study were not selected on the basis of increased cancer risk; therefore, they may have had low baseline levels of damage that were not amenable to further reduction by dietary change.

Objective: Rebeccamycin analog (NSC-655649) is an antibiotic with antitumor properties demonstrated in preclinical and phase I studies. We conducted a phase II trial to evaluate the efficacy and toxicity of this agent in patients with advanced renal cell cancer (RCC). Methods: Eligible patients had histologically or cytologically confirmed diagnosis of RCC that was either locally advanced unresectable, locally recurrent, or metastatic. Patients had to have measurable disease, no prior chemotherapy, life expectancy of greater than 12 weeks, an Eastern Cooperative Oncology Group (ECOG) performance status of 0–2, adequate-organ function, and be ≥18 years old. Patients were treated with NSC-655649 at a dose of 165 mg/m2 daily i.v. over 30–60 min for 5 days. Treatment was repeated every 21 days. Response was assessed every two courses. Results: Twenty-four patients were enrolled. There were sixteen males and eight females with a median age of 60.5 years (range 42–76). Nineteen were Caucasians, seventeen had prior nephrectomy, and thirteen had prior immunotherapy. The major toxicity was myelosuppression with grade 3 and 4 neutropenia in 38% of patients and anemia in 33% of patients. There were two partial responses (2/24, 8%) and 11 patients (46%) achieved stable disease (SD). The 6-month progression-free rate for patients with SD was 30%. Of the seventeen patients with progressive disease at registration, one had a PR and eight had SD. The overall median survival time for all 24 patients was 10.0 months (90% CI=5.2, 17.4 months). The 12-month survival rate was 39%, with 90% CI=(0.21, 0.58). Nine patients are still alive with survival times ranging from 3.8 to 24.2 months, at a median follow-up time of 11.9 months. Conclusion: Rebeccamycin analog (NSC-655649) is well tolerated and has modest antitumor activity in patients with advanced RCC.

Purpose: This phase II study was undertaken to determine the efficacy of adding infusional fluorouracil (FU) to the chemotherapy doublet of gemcitabine and cisplatin in patients with advanced pancreatic cancer. Patients and Methods: The eligibility criteria included histologically or cytologically confirmed adenocarcinoma of the pancreas that was either unresectable or metastatic. No prior gemcitabine therapy was allowed. Patients received a combination of gemcitabine 1,000 mg/m2 intravenously (IV) on days 1, 8, and 15; cisplatin 50 mg/m2 IV on days 1 and 15; and FU 175 mg/m2/d from days 1 to 15 by continuous IV infusion. Cycles were repeated every 28 days. Objective tumor response and toxicity were evaluated according to the World Health Organization criteria. Results: A total of 47 patients (median age, 57 years; males, 59%) were enrolled. Sixteen patients had locally advanced (LA) disease, and 31 patients had metastatic disease. A total of 183 cycles of chemotherapy were administered. In patients with metastatic disease (n = 31), the probability of survival at 6 and 12 months was 66% and 34%, respectively. Objective partial response or stable disease was observed in 26% (90% confidence interval [CI], 0.14 to 0.41) and 61% (90% CI, 0.45 to 0.74) of patients, respectively. In patients with LA disease (n = 16), there were three partial responses (19%; 90 CI, 0.07 to 0.39). One patient in this group was successfully resected after FU-based radiotherapy. The most common grade 3 to 4 toxicities were neutropenia (60%), thrombocytopenia (42%), and anemia (26%). Thirteen patients were hospitalized for treatment-related complications. Conclusion: The combination of gemcitabine, cisplatin, and infusional FU has significant activity in patients with advanced pancreatic cancer.

The spicamycin derivative KRN5500 was considered as a potential anti-cancer agent based on in vitro and preclinical studies. A Phase I study involving 24 cancer patients in whom tumors were refractory to all other conventional therapies was conducted to determine the dose limiting toxicity, maximum tolerated dose, effectiveness, and pharmacokinetic parameters of this drug administered by 1-h IV infusion daily for five consecutive days every 3 weeks. Using an accelerated dose titration strategy, 8.4 mg/m2/d × 5 days was the maximum administered dose. Severe gastrointestinal and hepatic toxicities were observed at doses at or above 4.3 mg/m2/d × 5. The recommended Phase II dose is 4.3 mg/m2/d × 5. The distribution of KRN5500 followed a two-compartment model, and clearance did not decrease significantly over the dose range 0.8–8.4 mg/m2/d × 5. No significant correlation was observed between plasma levels and toxicity. No tumor responses were observed among the 14 patients evaluable for response.

The cytochrome P-450 (CYP) and glutathione S-transferase (GST) enzyme systems may influence the biological effects of carcinogens, including estrogens. As such, these enzymes may predict the developmental risk of breast cancer, as well as be potential targets for chemoprevention. The purpose of this study was to compare the expression of GST-Pi and CYPs 1A1, 2B6, 2E1, and 3A4 in paired samples of normal and malignant breast tissue from patients with breast cancer and women undergoing reduction mammoplasty.Expression of CYPs 1A1, 2B6, 2E1, 3A4, and GST-Pi was quantified in breast tissue from 33 patients with breast cancer and in 17 women without history of cancer who underwent reduction mammoplasty. The expression of CYP 1A1, 2B6, 2E1, 3A4, and GST-Pi was quantified by immunoblotting.CYP 1A1, 2E1, and 3A4 expression was significantly lower (P0.05) in malignant tissue as compared with morphologically normal adjacent tissue. Conversely, GST-Pi expression was marginally lower in the normal tissue (P = 0.08). No significant difference in enzyme expression was seen between the tissue from reduction mammoplasty and normal tissue from breast cancer patients. There was a trend for higher expression of CYP 2B6 and GST-Pi in the estrogen receptor expressing tumors than those tumors without expression (P0.28).The expression of these enzymes was similar in morphologically normal breast tissue from patients with or without breast cancer. The expression of CYPs was down-regulated in the tumor tissue. The clinical significance of CYP alterations in breast cancer will need further characterization.

Objective: We examined plasma levels of carotenoids, tocopherols, and total antioxidant activity in women before and after dietary intervention to reduce fat and/or energy intakes. Dietary fat and energy may affect intake and bioavailability of carotenoids and tocopherols, and these micronutrient levels in turn can contribute to the antioxidant capacity of plasma. Methods: Women were randomized onto one of four diets for 12 wk: non-intervention, low fat (15% of energy from fat with maintenance of energy intake), low energy (25% energy reduction with maintenance of percentage of energy from fat), and combined low fat and low energy. Fasting plasma was available for analysis from a subset ( n = 41) of women enrolled in the study. Results: Levels of carotenoids and tocopherols did not change significantly over 12 wk on any diet arm, despite a modest but statistically significant increase in fruit and vegetable intake in the women following the low-fat diet (from 3.3 to 5.2 servings/d excluding potatoes). Levels of Trolox-equivalent antioxidant capacity (TEAC), total cholesterol, and two major plasma antioxidants (urate and bilirubin) also did not change significantly. Of the individual micronutrients measured, lycopene and lutein/zeaxanthin correlated most strongly with TEAC values, and the correlation with lycopene was statistically significant before intervention. Conclusions: The decreases in dietary fat and energy intakes in this study were quite large, but this did not appear to have detrimental effects on plasma micronutrient levels, nor did it appreciably affect plasma antioxidants. Because lycopene levels were significantly associated with plasma TEAC before intervention, interventions that increase levels of lycopene might be more likely to increase the antioxidant capacity of plasma.

5-Fluorouracil-based combination chemotherapy is commonly used in patients with advanced gastric cancer, but results with such therapy are fairly modest. Evaluation of newer agents is therefore required in this disease. Paclitaxel has shown promising activity as a single agent in gastric cancer. In vitro, paclitaxel exhibits sequence-dependent synergy with platinum compounds against gastric cancer. This study was conducted to evaluate the efficacy and toxicity of combination carboplatin and paclitaxel in patients with advanced gastric cancer. Twenty-seven patients with measurable or evaluable advanced gastric cancer were enrolled on the study from April 1996 to July 2000. Patients were treated with paclitaxel 200 mg/m intravenously during 3 hours followed by carboplatin at projected area under the curve 5 mg x h x ml (as per the Calvert formula). Twenty-six patients were assessable for toxicity, and 25 patients were assessable for objective response. Nine of the 27 enrolled patients had a major response for an objective response rate of 33% (95% CI 0.17-0.54) by intention-to-treat analysis. The median response duration was 4.9 months (95% CI 2.8-7.3), and median survival was 7.5 months. The 1-year survival rate was 23%. One hundred seventeen courses were administered with a median of four courses per patient administered. The major toxicity was neutropenia, with grade III to IV neutropenia observed in 9 patients (33%) and neutropenic fever in only 1 patient. Grade III peripheral neuropathy developed in two patients, and grade III myalgia and grade III fatigue developed in one patient each. There were no treatment-related deaths. The combination of carboplatin and paclitaxel is a highly tolerable, regimen with activity comparable to that of other regimens in advanced gastric cancer. This regimen needs to be further evaluated in combination with other agents and as a component of multimodality therapy in gastric cancer.

To compare the effects of low-fat, low-energy and combination low-fat/low-energy intervention on changes in six anthropometric measures in Caucasian and African-American free-living women.The effects of dietary counseling strategies for fat and/or energy reduction were examined on anthropometric measures in 86 pre-menopausal women, average BMI of 28 kg/m2, who participated in a 12-week intervention trial called the Women's Diet Study. The dietary goals were 15% of energy from fat and/or 25% reduction in energy intake, relative to reported baseline intake, using a 2 x 2 factorial design. Analysis of covariance models were constructed to evaluate changes in anthropometric measures over the 12 weeks of study.The biggest difference by race was in women who were relatively heavier at baseline, in which case African-American women lost significantly less weight but decreased their waist:hip ratio to a significantly greater extent than Caucasian women. With regard to the effects of diet arm, weight loss varied depending on baseline weight, and in women with higher baseline weights, the combination low-fat/low-energy diet resulted in the most weight loss (6.7 kg, p0.05). Decreases in the other anthropometric measures at week 12 were more uniform across diet arms and did not depend on baseline values. After controlling for previous weight history and race, the decreases in BMI, percent body fat and waist circumference after 12 weeks were statistically equivalent with the low-fat, low-energy or combination low-fat/low-energy diets. The relatively greater decreases in percent body fat and waist circumference with the combination diet versus the low-fat or low-energy diets were not statistically significant.The low-fat, low-energy and combination diets all resulted in similar and statistically significant decreases in BMI, percent body fat and waist circumference over 12 weeks of intervention. The extent of weight loss, however, varied depending on baseline weight, and the combination diet was the only intervention to result in significant weight loss for women who were heavier at baseline. This indicates that, although there may be an advantage for reducing dietary fat in initially heavier women, any of these counseling strategies could be effective for improving anthropometric predictors of health risks associated with overweight status. This is useful since flexibility in dietary choices may facilitate adherence to dietary counseling in some individuals.

5060 Background: Neuroendocrine (N) differentiation is a resistance mechanism to chemotherapy in mCRPC. Somatostatin (sst) receptor analogs demonstrate efficacy in N tumors. P is a multitargeted ss...

No. 446 Lung cryoablation offers procedure flexibility P. Littrup, B.F. Baigorri, H.D. Aoun, B. Adam, E. N. Fletcher, M. Krycia, L. Heilbrun; Karmanos Cancer Institute, Detroit, MI; Wayne State University, Detroit, MI Purpose: To assess complication and recurrence factors for percutaneous thoracic cryoablation. Tumor and ablation size, complications, location and vessel proximity were assessed for patients with primary thoracic and metastatic tumors. This included subset comparisons between a prospective cohort which defined PET-CT follow up, with the remaining nonprospective cohort. Materials and Methods: CT and/or CT-US fluoroscopicguided percutaneous cryoablation was used in a total of 177 procedures on 303 tumors (120 primary, 183 metastatic tumors) in 132 patients, noting tumor and ablation volumes, recurrences, complications, and tumor type. Of these, 41 patients, 51 procedures, and 65 tumors were on the prospective protocol. Primary thoracic included all lung cancer types (n1⁄470) and pleural tumors (n1⁄45). Complications were graded by Common Terminology of Complications and Adverse Events (CTCAE v4.0). Hydrodissection and esophageal warming balloon were used for tissue separation as needed (20 and 9 respectively). A minimum of 2 cryoprobes were used on all patients, probe number was generally based on a formula of tumor diameter plus one. Results: All patients required only conscious sedation. No significant difference was noted for patient criteria or their outcomes between the prospective and non-prospective cohort subsets. Data were therefore grouped for greater power. Overall tumor and ablation median size was 2.2cm and 4.2cm, respectively. Major complication rates were significantly lower in tumors r3 cm as opposed to Z3cm, 1.5% (2/134) vs. 11.8% (9/76) (po0.005). Total major complication rates were low at 6.2% (11/177). Low total tumor recurrence rates of 6.9% (21/ 303) were noted as progression in 2.6% (8/303) and satellite recurrence in 4.3% (13/303). Conclusion: CT guided percutaneous cryoablation in the lung provides a low morbidity alternative for complex patients, particularly for pleural/chest wall and more central tumors. Complication rates are significantly lower for tumors o3cm. Freezing well into chest wall for pleural tumors showed good healing with no complications. Recurrence rates were not affected by tumor size. Appropriate case selection allows for low recurrence and complication rates. Educational Exhibit Abstract No. 447 CT-guided autologous blood patch for the post transthoracic lung intervention pneumothorax S. Bansal, F. Genshaft, A. Abtin, R. Gutierrez; Suh Thoracic Imaging and Intervention, Department of Radiological Sciences, David Geffen School of Medicine at UCLA, Los Angeles, CA Learning Objectives: The objective of this exhibit is to present the technique for CT-directed autologous blood patching in the treatment of persistent pneumothorax following percutaneous lung biopsy and ablation. Background: Pneumothorax is a common complication of percutaneous thoracic interventions, including both lung biopsy and ablation, complicating up to 50% of procedures. Small pnemothoraces can often be treated conservatively, however, pneumothoraces with persistent air leak can be difficult and frustrating to treat, and at times require thoracic surgery for definitive management. As described in the surgical literature, the mechanism of blood patching is thought to involve clot formation to seal persistent air leaks from the lung parenchyma. We use an illustrative case-based approach to present a twocatheter image-guided blood patch technique to seal the persistent air leak, drain intrapleural air, and oppose the pleural surfaces, effectively treating a bronchopleural fistula. Clinical Findings/Procedure Details: 1. Clinical assessment and management of post-intervention pneumothorax. 2. Treatment options for a persistent bronchopleural fistula failing conservative catheter drainage. 3. Two-catheter technique for CT-guided blood patch placement. 3a. Percutaneous autologous blood patch. 3b. Pneumothorax aspiration. 3c. Thoracostomy catheter management. 4. Post-procedural observation and hospital management Conclusion and/or Teaching Points: Pneumothorax is the most common post-procedural complication following transthoracic lung intervention. Persistent air leaks are associated with significant morbidity and increased length of hospital stay. Given the increasing use of percutaneous transthoracic lung biopsy in the diagnostic algorithm and thermal lung ablation in the treatment of malignancy, it is important for interventionalists to be comfortable treating this complication. We have had success with CT-guided autologous pleural blood patching for management of patients with persistent air leaks following lung intervention. Pleural blood patching appears to accelerate the healing of the bronchopleural fistula, allowing early catheter removal, providing a valuable treatment option with minimal side effects. Educational Exhibit Abstract No. 448 Diagnosis and treatment of isolated splenic vein thrombosis S. Wong, M. Cristescu, J.F. Angle, Z.J. Haskal, J. Stone, L.R. Wilkins, S.S. Sabri; UVA, Charlottesville, VA; Univeristy of Wisconsin, Madison, WI Learning Objectives: Learning Objectives1. Understand the pathophysiology and clinical presentation of isolated splenic vein thrombosis (ISVT)2. Review the role of imaging in the management of ISVT3. Review the treatment options for ISVT Background: ISVT is a rare condition that results in left sided portal shunting. It is associated with several disease processes but most commonly with pancreatitis. Clinical symptoms are often vague and physical exam findings often misleading of the diagnosis, requiring the clinician to hold a high index of suspicion. Although most patients are asymptomatic, variceal bleeding is a potentially serious complication. Currently only a handful of treatment options are available to symptomatic patients. Clinical Findings/Procedure Details: Patients with symptomatic ISVT often have a past medical history of chronic pancreatitis and experience vague abdominal pain or variceal bleeding. Splenomegaly is not a reliable physical exam finding. 20% of symptomatic patients do not have esophogastric varices due to an anatomic variant of the cardiac vein. CT or MRI are Posters and Exhibits ’ JVIR S194 P os te rs an d Ex hi bi ts

442 Background: Clinical utility and ability to prognosticate using FDG-PET scan in metastatic colorectal cancer (mCRC) has not been well studied in previous literature. We hypothesized that standardized uptake value (SUV) in post treatment FDG-PET predicts clinical outcomes. Methods: We retrospectively reviewed mCRC patients who had pre and post treatment FDG - PET scans during 2005-2012 at Karmanos Cancer Institute. Primary end points of time to progression (TTP) and overall survival (OS) were compared between the two groups (follow up SUV = 0 vs. SUV > 0) using log-rank testing. Results: Study population consisted of 44 patients (median age of 58.1 years). Forty (91%) of these patients were treated first line and 34 (77 %) received an Oxaliplatin based regimen. Forty-one patients had both pre- and a post-treatment SUV. The 12 patients with post treatment SUV= 0 and the 29 patients with post-treatment SUV > 0 had median OS of 42 and 25.2 months, respectively; they had median TTP of 8.2 and 6.9 months, respectively. The patients with post treatment SUV= 0 had significantly longer OS (p = 0.0310) and weakly significantly longer TTP (p = 0.0628) compared to the patients with post-treatment SUV > 0. Conclusions: Systemic therapy decreased SUV on follow-up PET scans in the study population. Absence of FDG uptake on follow up PET scan was associated with longer OS and was suggestive of longer TTP in patients with metastatic colorectal cancer.

Summary Background Satraplatin is an oral platinum compound that has demonstrated efficacy and tolerability in prostate cancer. Preclinical synergy between bevacizumab and platinum has been noted. Methods Docetaxel-pretreated metastatic castrate-resistant prostate cancer patients with disease progression were eligible. Satraplatin 80 mg/m 2 orally on days 1 to 5, prednisone 5 mg twice daily, and bevacizumab 10 mg/kg on day 1, and 15 mg/kg on day 15 were administered in 35-day cycles. Results Thirty one patients were enrolled. Grade 3 or 4 toxicities were pulmonary embolism in 2 patients and thrombocytopenia in 1 patient. 31% of the patients had a≥30% decline in prostate-specific antigen. Median time to progression was 7.0 months (90% confidence interval [CI] 4.7–8.5 mo) and median overall survival was 11.2 months (90% CI 9.1–16.4 mo). Polymorphism in the excision repair cross-complementation-1 (ERCC-1) gene was associated with time to progression (hazard ratio = 1.91). A circulating tumor cell count≥5was moderately prognostic of overall survival (hazard ratio = 1.49) as compared with CTC Conclusions The combination was tolerable, and revealed promising efficacy in metastatic castrate-resistant prostate cancer. ERCC1 genotype maybe predictive of clinical benefit with platinum-based therapy in metastatic prostate cancer.

Antioxidant micronutrients found in fruits and vegetables have been shown in numerous studies to be protective against cancer. There is limited information on the relationship between blood antioxidant micronutrient levels and cancer among ethnic minorities. We conducted a pilot case-control study to evaluate the potential for accrual to a study of the association of plasma levels of beta-carotene, retinol, lycopene, alpha-tocopherol, and gamma-tocopherol with breast cancer risk among African American and Caucasian women seen at a large university medical center in Detroit. Cases included women with newly diagnosed invasive breast cancer who had not yet had any cancer-related therapy and who were age-matched to controls within 5 years. Plasma levels of micronutrients were analyzed by high-pressure liquid chromatography. Compared to the expected accrual based on cancer registry data, only 26% (11/42) of African American women with breast cancer enrolled, while 100% (16/16) of Caucasian cases enrolled. Control women were quickly accrued with only a 6% refusal rate. Among African American women, there was a weak inverse association between plasma lycopene levels and breast cancer risk, with a mean level of 0.17mgr;mol/L (SD = 0.18) among cases, and 0.24mgr;mol/L (SD = 0.18) among controls (p = 0.09). There was a weak direct association between plasma retinol levels and breast cancer risk among African American women, with a mean retinol level of 2.37mgr;mol/L (SD = 0.73) among cases and 1.98mgr;mol/L (SD = 0.49) among controls (p = 0.132). The interaction effect of race and lycopene was statistically significant (p = 0.048). Among the lowest lycopene tertile, the risk of breast cancer among Caucasian women was 0.76 and the risk of breast cancer among African American women was 2.29, although these odds ratios were not statistically significant. Our recruitment efforts were largely successful among Caucasian cases and controls, and African American controls, but were unsuccessful among African American cases. The results suggest a possible relationship between plasma lycopene level and breast cancer among African American women, but these results should be confirmed by a larger, more definitive study.

A highly active and broadly active thioxanthone has been identified: N-[[1-[[2-(Diethylamino)ethyl]amino]-7-methoxy-9-oxo-9H-thioxanthen++ +-4-yl] methylformamide (SR271425, BCN326862, WIN71425). In preclinical testing against a variety of subcutaneously growing solid tumors, the following %T/C and Log10 tumor cell kill (LK) values were obtained: Panc-03 T/C = 0, 5/5 cures; Colon-38 (adv. stage) T/C = 0, 3/5 cures, 4.9 LK; Mam-16/C T/C = 0, 3.5 LK; Mam-17/0 T/C = 0, 2.8 LK; Colon-26 T/C = 0, 1/5 cures, 3.2 LK; Colon-51 T/C = 0, 2.7 LK; Panc-02 T/C = 0, 3.1 LK; B16 Melanoma T/C = 13%, 4.0 LK; Squamous Lung-LC12 (adv. stage) T/C = 14%, 4.9 LK; BG-1 human ovarian T/C = 16%, 1.3 LK; WSU-Brl human breast T/C = 25%, 0.8 LK. The agent was modestly active against doxorubicin (Adr)-resistant solid tumors: Mam-17/AdrT/C =23%, 0.8 LK; and Mam-16/C/Adr T/C = 25%, 1.0 LK, but retained substantial activity against a taxol-resistant tumor: Mam-16/C/taxol T/C = 3%, 2.4 LK. SR271425 was highly active against IV implanted leukemias, L1210 6.3 LK and AML1498 5.3 LK. The agent was equally active both by the IV and oral routes of administration, although requiring approximately 30% higher dose by the oral route. Based on its preclinical antitumor profile, it may be appropriate to evaluate SR271425 in clinical trials.

A series of quinoxaline analogs of the herbicide Assure® was found to have selective cytotoxicity for solid tumors of mice in a disk-diffusion-soft-agar-colony-formation-assay compared to L1210 leukemia. Four agents without selective cytotoxicity and 14 agents with selective cytotoxicity were evaluated in vivo for activity against a solid tumor. The four agents without selective cytotoxicity in the disk-assay were inactive in vivo (T/C > 42%). Thirteen of the fourteen agents with selectivity in the disk-assay were active in vivo (T/C < 42%). Five of the agents had curative activity. These five agents had a halogen (F, Cl, Br) in the 7-position (whereas Assure® had a Cl in the 6 position). All agents with curative activity were either a carboxylic acid, or a derivative thereof, whereas Assure® is the ethyl ester of the carboxylic acid. All other structural features were identical between Assure® and the curative agents. Assure® had no selective cytotoxicity for solid tumors in the disk-assay, and was devoid of antitumor activity. The analog XK469 is in clinical development.

Purpose: Pyrazoloacridine (PZA) is an acridine derivative selected for clinical development because of broad preclinical antitumor activity and solid tumor selectivity. Phase I evaluations with PZA have demonstrated predictable toxicity and suggested clinical efficacy. A phase II trial in patients with previously untreated advanced pancreatic cancer was conducted. Methods: PZA was administered at a dose of 750 mg/m2 intravenously over 3 hours every 21 days. Seventeen patients were treated receiving a total of 46 courses of PZA. Results: Of the 15 patients evaluable for response, no responses were observed (0% response rate, 95% confidence interval 0–22%). Major toxicities directly attributable to PZA included moderate neutropenia and mild neurotoxicity. Conclusion: PZA at this dose and schedule of administration was inactive in patients with pancreatic carcinoma.

Will early detection impact on stage of disease and recurrence of prostate cancer in a high risk population? We initiated a community based study to educate and recruit African American men for early diagnosis of prostate cancer, that is the Detroit Education and Early Detection (DEED) study. Our objective was to evaluate our recruitment process for this target population, examine the percentage of organ confined prostate cancer in men undergoing radical prostatectomy and measure recurrence biochemically.A community based study from February 1993 to February 1995 through the African American churches in metropolitan Detroit was initiated. We compared the early detection group treated with radical prostatectomy to the population presenting to our urological clinic during the same period. We tested and followed 1,105 African American men using the prostate specific antigen blood test.Pathologically organ confined prostate cancer was diagnosed in 11 of 17 men (65%) who underwent radical prostatectomy in the DEED project. Within the clinic population 35% of the African American men were diagnosed with pathologically organ confined prostate cancer. The difference between the 2 populations was statistically significant (p = 0.033). Disease recurred in 1 of 15 (7%) and 39 of 157 (25%) men in the DEED and clinic populations, respectively (p = 0.97).We demonstrated our ability to recruit African American men into a prostate cancer early detection program. We diagnosed early but clinically significant prostate cancers among African American men with characteristics similar to prostate cancers diagnosed in other early detection studies in which the overwhelming majority of men were white.

A randomized intervention trial of dietary fat reduction to 15% of total calories was initiated in 1987 for women at high risk for breast cancer to determine the feasibility of recruiting and maintaining them on a low-fat diet. The study has enrolled 194 women between the ages of 18 and 67 years who met at least one of three eligibility criteria: 1) a first-degree relative with breast cancer, 2) a P2 or DY Wolfe mammographic pattern, and 3) a prior breast biopsy demonstrating epithelial hyperplasia with or without atypia. Eligible women must also have had diets that containedor = 30% of calories from fat at entry. Women were randomized to a nonintervention usual diet vs. a 15% low-fat diet. Recruitment was sought through physicians, personal mailings, breast cancer patients, and the news media. Two study sites participated: a large urban hospital affiliated with a university medical center and a community oncology private practice. The results from both institutions were similar and demonstrated that a low-fat dietary plan could be effectively conducted in private as well as academic settings with recruitment tailored to the community where the trial is being conducted. Reduction in dietary fat intake was maximal during the first three months of the dietary intervention and remained stable throughout 12 months of follow-up. Reductions in total calories, weight loss, and percent body fat were minimal. The nonintervention group experienced no major change in their diet. We conclude that it is feasible to recruit women who are at high risk for breast cancer into a dietary intervention trial and with sufficient dietary counseling and motivation on the part of participants, reduction in dietary fat intake can be achieved and maintained. More in-depth analyses of these data will be presented in subsequent reports.

5041 Background: A phase II clinical trial was conducted of the combination of C and E due to the synergy noted. Methods: The primary endpoint was time to progression (TTP). Intravenous C at a target AUC of 5 on day 1, and oral E 5mg once daily and P 5mg twice daily were administered in 21 day cycles. PSA was assessed every 21 days and radiologic response was assessed every 3 cycles. Secondary endpoints included overall survival (OS), the correlation of TTP with phosphorylated (p) mTOR, pAKT, p70S6, and circulating tumor cells (CTC). A 1-stage study design assumed: a reference median TTP = 1.5 months; 1-sided alpha = 0.15; and power = 0.90, requiring 26 patients (pts). Results: 26 pts enrolled; median age 69 years (range 54-86) ;8 African American and 18 Caucasians. Median pretherapy PSA was 190 ng/ml (range 13 - 2174). 18 pts (69%) each had bone pain and Gleason score > 8. 125 cycles have been administered; median 3 cycles (range 1 - 16). Predominant grade 3 or 4 toxicities were thrombocytopenia in 8 pts, pulmonary embolism in 2 and neutropenia in 3. No treatment related deaths occurred. 4 (15%) had a > 30% PSA decline and 1 had a >90% PSA decline. 8/19 pts had stable disease but no objective responses in MD. The median TTP and OS were 2.5 months (90% CI: 1.8 - 4.3), and 12.5 months (90% CI: 6.7 - 16.1), respectively. Median area under curves were 5.9 (range, 4.3 – 11.0) and 4.5 (range, 4.1 – 7.1) mg/mL*min with C given alone and in combination with E, respectively. E did not influence pharmacokinetics of C. Median baseline CTC (n=18) was 30 (range 0-2372). 5/18 pts had favorable CTC (CTC18 weeks had reduction in post-therapy CTC with a median decrease of 63% (range 11%-100%). Lack of IHC staining for pAKT was noted in 2/2 pts on therapy for > 30 weeks vs increased expression was noted in 8/8 pts on therapy for < 9 weeks. Testing for TSC1 mutation is planned and will be reported. Conclusions: The combination was tolerable but revealed modest clinical efficacy. Biomarker evaluations such as pAKT may help identify a subset likely to benefit from mTOR inhibitor strategy in mCRPC. Clinical trial information: NCT01051570.

5085 Background: Hsp90 is a molecular chaperone required for the proper folding and activation of numerous client proteins and is critical to cell survival and proliferation. Ganetespib (G), a synthetic small molecule that binds to the ATP pocket in the N-terminus of Hsp90 causes the degradation of cellular proteins and ultimately death of cancer cells dependent on these proteins. G displays potent anticancer activity in prostate cancer cells. Methods: Eligible patients were ≥ 18 yrs old with ECOG performance status ≤ 2. Adequate hepatic, renal, and bone marrow function was required. Patients were treated with G 200 mg/m2 intravenously once weekly for 3 consecutive weeks followed by 1 off-week. The primary objective was to evaluate the 6-month progression-free survival (PFS). Secondary endpoints included overall safety and tolerability of G and overall survival. Exploratory markers including maspin, cytokeratins 8 and 18, and HDAC1 were obtained pre, during, and post G treatment. We sought ≥ 4 patients with ≥ 6 months PFS (a success) in Stage 1 of a 2-stage near-optimal Simon design. Results: 18 patients were enrolled at 4 institutions. The patients’ median age was 68 (range 51-82), 13 were Caucasian, 4 were African American, and 1 was Asian. Median PSA was 210.7 ng/mL (range 25.9 – 3,489 ng/mL). One patient never started therapy. Among the 17 treated patients, the median number of cycles was 2 (range 1-5). The most frequent Grade 3 toxicities were diarrhea (3 patients), fatigue (3), and dehydration (3). Prior therapy included abiraterone (8), sipuleucel-T (4), and other targeted therapy (4). With < 4 successes in Stage 1, the trial was terminated early. Median PFS was 1.9 months (90% CI: 1.7 – 2.7 months); median OS was 10.2 months (90% CI: 2.3 – 18.3 months). Exploratory markers have been evaluated and will be presented. Conclusions: Our study represents the first clinical trial of G in treating mCRPC patients. As a single agent therapy, G did not prolong PFS. Combination therapy is a consideration to improve therapeutic efficacy. Clinical trial information: NCT01270880.

6045 Background: SAHA is a small molecule inhibitor of histone deacetylases (HDAC). Based on confirmed responses noted in 2 ACC patients treated with SAHA at our institution; we initiated this phase II trial. Methods: Patients with LA, recurrent or metastatic ACC and adequate organ function were eligible. Any prior number of chemotherapy regimens was allowed. Tumor tissue and blood were collected for correlative studies. SAHA was administered orally 400 mg daily for 28 days. The primary objective was to evaluate response rate (RR). Secondary endpoints included: time to tumor response (TTR); response duration (RD); progression-free survival (PFS); overall survival (OS); and adverse events (AE). 29 evaluable patients were needed in a Simon 2-stage optimal design to distinguish RR of at most 5% vs. at least 20%, with alpha = 0.15 and power = 0.90. Results: 30 evaluable patients (overenrolled by 1) were accrued from at 5 different centers between August 2010 and June 2012. Median follow up among the censored patients is 3.8 months for PFS, and 7.2 months for OS. 19 patients were female, 24 were Caucasian, median age was 53 years (range 21-73); 21 patients had a performance status of 1; 20 patients were chemo-naïve. The median number of cycles completed was 5 (range 1-27+). Lymphopenia (n=5), hypertension (n=3), oral pain (n=2), thromboembolic events (n=2) and fatigue (n=2) were the only grade 3 AEs that occurred in more than 1 patient. 5 patients were dose reduced due to AEs. At the time of data cut-off, 9/30/2012, 1 patient with lung metastasis had a PR, with RD of 11.2+ months. TTR was 7.7 months. SD was the best response in 25 patients. Median PFS is 12.7 months (90% confidence interval lower limit: 3.7 months). Median OS has not been reached. There were 6 patients with PFS > 1 year. The correlative studies results will be presented. Conclusions: SAHA shows promise in the treatment of ACC. We are awaiting genomic analysis of the tumors to study the basis of clinical benefit (PR and SD) of SAHA in ACC. Future drug combination studies in ACC are being considered focusing on the epigenetic mechanism of SAHA. Supported by 5U01CA062487/19. Clinical trial information: NCT01175980.

38 Background: Cediranib + Doc were shown in a phase I trial to be feasible, with early evidence for efficacy in mCRPC patients (pts). A multi-center randomized Phase II screening trial of Cediranib+Doc+Pred (Arm A) vs. Doc+Pred (Arm B) was initiated in mCRPC pts. Methods: mCRPC pts with no prior chemotherapy were eligible. All received Doc at 75 mg/m2IV q 3 weeks and Pred 5 mg po bid. Cycle length was 21 days. Arm A pts also received Cediranib at 30 mg po daily. The primary endpoint was 6-month progression-free survival (PFS). We hypothesized 6-month PFS rates of 0.70 on Arm A and 0.50 on Arm B. Secondary endpoints included toxicity, PSA response, and RECIST response. Here we report the toxicity and PSA response data. Results: Of 58 pts enrolled, 57 were treated (29 Arm A, 28 Arm B). Median age = 68 years (range 51-84); 33% African-American, 63% Caucasian; median baseline PSA136 ng/mL (range 0.12 – 3,650); 57% had Gleason grade 8-10; and 40% had both bone/ visceral disease. Median nadir PSA was 10.5 ng/mL (range 0.10 – 1,484) in Arm A, and 25.9 ng/mL (range 0.00 – 2,076) in Arm B. Median number of cycles (range): Arm A = 9 (0 - 31); Arm B = 6.5 (1-39 ). 68% pts in Arm A required a dose reduction of at least one Cediranib dose level. Doc dose was reduced in 45% pts on Arm A and in 18% pts on Arm B. Primary grade 4 toxicity was neutropenia (11 [38%] pts A; 5 [18%] pts B), which prompted Cediranib dose reduction. Grade 3 toxicities included fatigue (9 (31%) A, 1 (4%) B); hypertension (8 (28%) A, 1 (4%) B); anemia (6 (21%) A, 2 (7%) B); neutropenia (4 (14%) A; 6 (21%) B); diarrhea (3 (10%) A, 1 (4%) B); deep vein thrombosis (2 (7%) A, 1 (4%) B); and pulmonary embolism (1 (3%) A). An amendment reduced Cediranib to 20 mg daily dose. PSA decrease ≥ 90% at nadir occurred for 13/29 = 45% on Arm A, and 6/28 = 21% on Arm B. PSA decrease ≥ 50% at nadir occurred for 19/29 = 66% on Arm A, and 17/28 = 61% on Arm B. Clinical partial response rates were 8/15 = 53% on Arm A and 3/9 = 33% on Arm B. Conclusions: The addition of Cediranib to Doc/Pred has increased toxicity but may be associated with higher rates of PSA response and clinical response in mCRPC patients. PFS results will be reported. Clinical trial information: NCT00527124.

156 Background: Platinum based therapies have demonstrated efficacy in DP mCRPC. EVE demonstrated preclinical efficacy in chemotherapy resistant prostate cancer models. Clinical synergy was noted between C and EVE, hence a phase II trial of the combination was conducted. Methods: Primary endpoint was time to progression (TTP). Progression was defined per RECIST criteria for measurable disease (MD), or skeletal event, or > 2 new areas of bone metastases. DP mCRPC patients with adequate renal and liver function, and performance status of 0 or 1 were eligible. Intravenous C at target AUC of 5 on day 1, and oral EVE 5mg once daily and P 5mg twice daily were administered in 21 day cycles. PSA was assessed every 21 days and radiologic response was assessed every 3 cycles. Secondary endpoints included overall survival (OS), correlation of TTP and PSA response, with markers such as phopho mTOR, pAKT, p70S6 and circulating tumor cells (CTC). Results: 26 patients (pts) enrolled, including 8 African Americans, and accrual is complete. Median age was 69 years (range 54-86). Median pretherapy PSA was 190 ng/ml (range 13 - 2174). 18 pts (69%) had bone pain. Gleason score was > 8 in 18 pts. 19 pts had measurable disease of which 15 had MD progression, 18 had bone scan progression, and 2 had PSA-only progression. 124 cycles have been administered; median 3 cycles (range 1 - 16). The predominant grade 3 or 4 toxicities were thrombocytopenia in 8 pts, pulmonary embolism in 2 and neutropenia in 3. No treatment related deaths occurred. Of 26 pts who are response evaluable, 4 (15%) had a > 30% PSA decline and 1 had a >90% PSA decline. Of 19 pts with MD, 8 had stable disease and no objective responses were observed. The median TTP and OS were 2.5 months (90% CI: 1.8 - 4.3), and 12.5 months (90% CI: 6.7 - 16.1), respectively. Correlative studies including pharmacokinetic and pharmacodynamic evaluations are ongoing, and will be reported. Conclusions: The combination was tolerable but revealed modest clinical efficacy. Biomarker evaluation may help identify a subset likely to benefit from mTOR inhibition strategy in mCRPC. Clinical trial information: NCT01051570.

Aims: To study the prevalence, extent and evolution of high-grade prostatic intraepithelial neoplasia (HGPIN) in African-American and Caucasian men of a wide age range in order to help clarify the role this lesion may play in the racial differences of prostate cancer. Methods: The lesion was documented in step-sectioned, entirely submitted prostates of two study populations; 525 autopsied men who died of trauma and 1,000 patients who had retropubic radical prostatectomy for clinically localized carcinoma of the prostate. Results: We found that HGPIN starts in young individuals and increases progressively with advancing age in both races but is more prevalent in African- Americans. Additionally, the more extensive form of HGPIN with multifocal or diffuse involvement of the gland appears at a younger age in African-Americans. Conclusions: Microscopic foci of HGPIN can be documented in males in the 3rd and 4th decades with the lesion becoming more extensive in older men. The finding that HGPIN is both more prevalent and its more diffuse form appears earlier in African-Americans, indicates a potentially important role for this lesion in the race-related discrepancies associated with this disease.

The in vitro testing of antitumor drugs involves the use of mouse and human tumor cells. In particular, there is interest in developing agents active against human solid tumors. We examined several biochemical parameters that may contribute to the differential sensitivity of the cell lines used in our laboratory to the toxic effects of antitumor compounds. The tumor cell lines examined were of mouse (colon 38, L1210 leukemia, and C1498 leukemia) and human origin (CEM leukemia, CX1 colon, H116 colon, HCT8 colon and H125 lung). Quinone reductase activity was markedly different between leukemia and solid-tumor cell lines of either mouse or human origin, with increased activity being observed in the solid-tumor cell lines relative to the leukemia lines. GSH transferase activity also was generally increased in solid-tumor relative to leukemia cell lines. Superoxide dismutase activity and thiol levels were similar in leukemia and solid-tumor cell lines, except that thiol levels were very low in colon 38. Mouse cell lines from in vitro passage had somewhat higher activity of superoxide dismutase and thiol levels than did cells maintained in vivo, indicating relatively increased antioxidant defenses. The toxicity of 2,3-dimethoxy-1,4-naphthoquinone, a model quinone that exerts its toxic effects via production of reactive oxygen species, was significantly lower in mouse lines maintained in vitro than in those tested in vivo, whereas the toxicity of another quinone, menadione, was just slightly lower. Quinone reductase activity, GSH transferase activity, and thiol levels were significantly higher in the human lines than in the mouse lines. Accordingly, the toxicity of both quinones tended to be lower in the human lines than in the mouse lines.

Background Patients with metastatic head and neck squamous cell carcinoma (HNSCC) have a poor prognosis, limited treatment options, and median survival of 6 to 9 months. Paclitaxel (TAX) and gemcitabine (GEM) have both shown activity in HNSCC. The optimal combination, dosing, and scheduling of both drugs is, however, unknown. Thus, we investigated the efficacy and safety of biweekly dose-intense GEM/TAX in patients with recurrent locoregional or metastatic HNSCC. Methods An open-label, single-institution, single-arm, phase II study was conducted for patients (pts) who were previously treated with no more than two cytotoxic regimens. The pts received paclitaxel (150 mg/m2IV) and gemcitabine (3000 mg/m2 IV) on day 1 and 15. The treatments were repeated every 28 days (one cycle), until disease progression or unacceptable toxicity. The primary end point was response rate. RECIST-defined response was evaluated every 2 cycles (8 wks) and toxicities were evaluated after each cycle (4 wks). Results Fifty-five pts were enrolled into the trial (M: F 42:13, median age (range): 56 years (23-84), performance status 0-2, 48 pts had previous radiation therapy and 39 had previous surgery), 41 of whom were response evaluable. Of these 41 patients, two pts (4.8%) achieved complete response (CR) and 17 (41.4%) demonstrated a partial response (PR). Thus, the overall response rate was 46%. Thirteen pts (31.7%) had stable disease (SD), resulting in tumor control in 32 of 41 pts (78% with CR, PR or SD), whereas 9 pts (21.1%) had disease progression (PD). Among those pts who achieved objective response or stable disease, the median response duration was 5 months and median time to progression was 4 months. Median overall survival (OS) was 15 months. Myelosuppression was the most common adverse event. Grade 3-4 neutropenia and anemia were observed in 5 (12%) and 11 (26%) pts, respectively. Eight (19%) pts developed grade 3 infection. None of these pts, however, had febrile neutropenia and there were no treatment-related deaths. Conclusions The combination of biweekly dose intense GEM/TAX was tolerable and active regimen in patients with recurrent locoregional or metastatic HNSCC. Our findings warrant further investigation in a larger patient population. Disclosure All authors have declared no conflicts of interest.

5534 Notice of Retraction: "Phase II study of biweekly dose-intense docetaxel plus gemcitabine (GEM/DOC) in patients with recurrent locoregional or metastatic head and neck squamous cell carcinoma." ASCO's Confidentiality Policy requires that abstracts be considered confidential and embargoed from the time of submission until the findings have been publicly released in conjunction with the ASCO Annual Meeting. Abstract 5534, published in the 2012 Annual Meeting Proceedings Part I, a supplement to the Journal of Clinical Oncology, violated this policy and has been retracted from publication and presentation at the 2012 ASCO Annual Meeting. Background: Patients with metastatic head and neck squamous cell carcinoma (HNSCC) have a poor prognosis, limited treatment options, and median survival of 6 to 9 months. Docetaxel and gemcitabine have both shown activity in HNSCC. The optimal combination, dosing, and scheduling of both drugs is, however, unknown. Thus, we investigated the efficacy and safety of biweekly dose intense GEM/DOC in patients with recurrent locoregional or metastatic HNSCC. Methods: An open-label, single-institution, single-arm, phase II study was conducted for patients who were previously treated with no more than two cytotoxic regimens. The patients received docetaxel (60 mg/m2IV) and gemcitabine (3000 mg/m2 IV) on day 1. The treatments were repeated every 14 days (one cycle), until disease progression or unacceptable toxicity. The primary end point was response rate. RECIST-defined response was evaluated every 4 cycles and toxicities were evaluated at each cycle. Results: A total of 36 patients were enrolled (M:F 26:10; median age (range), 60 years (46-79); performance status 0-1) , 29 of whom were response-evaluable. The patients received a median of 4 cycles (range 0-24). Of these 29 patients, none achieved complete response (CR) and 6 demonstrated a partial response (PR). Thus, the overall response rate was 21% (95% confidence interval [CI], 0.10 – 0.38). Ten patients had stable disease (SD), resulting in tumor control (CR or PR or SD) in 16 of 29 patients (55%), whereas 13 patients (45%) had disease progression. The median response duration was 3.2 months (80% CI: 2.0 – 6.1 months). For all 36 patients, the median overall survival was 4.2 months (95% CI: 2.4 – 7.0 months). Myelosuppression was the most common adverse event. Grade 3-4 neutropenia and anemia were observed in 10 (30%) and 13 (39%) patients, respectively. None of these patients, however, had febrile neutropenia or bleeding events, and there were no treatment-related deaths. Conclusions: The combination of biweekly dose intense GEM/DOC was tolerable and active regimen in patients with recurrent locoregional or metastatic HNSCC. Our findings warrant further investigation in a larger patient population.

e15081 Background: Resection of metastasis in renal cancer (mRCC) yields 5-year survival rates of 20% to 30%. Yet most patients (pts) poor surgical candidates or have unresectable disease. Percutaneous cryoablation is a minimally invasive procedure employed for local therapy of metastatic foci, but its role in inducing long term remissions in combination with systemic therapy is unknown. Methods: Following IRB approval and HIPAA compliance, a retrospective review of records for pts receiving cryotherapy and systemic targeted therapy from 2004-2007 at Karmanos Cancer Institute was performed to enable a minimum follow up of 3 years. OS was estimated using the Kaplan-Meier methodology. Results: Of the 27 patients identified 85% were male, 88% Caucasian, and 77% had clear cell histology. 2 pts had sarcomatoid mRCC. All patients had prior nephrectomy. 24 patients had multiple metastatic sites: 9 with 2 sites, 8 with 3 sites and 7 with ≥ 4 sites. 15%, 52%, 11%, and 22% had favorable, intermediate, poor, and indeterminate MSKCC risk score, respectively. Systemic therapy was interrupted for at least 1 week prior to and after cryotherapy and resumed after complete healing. One major complication of nerve paresis was noted from lumbar paravertebral mass cryotherapy. Median number of ablations was 2 (range 1-4). Of the 60 total ablations 13%, 28%, 27%, 17%, and 15% were renal bed, visceral, soft tissue, bone, and lymph nodes respectively. The median OS from diagnosis of metastatic disease was 49.6 months (95% confidence interval: 21.2- 54.9 months).The 2 yr OS rate is 90% (95% CI: 78-100%) and 5 yr OS rate is 68% (49-87%). 8 patients (30%) are alive to date. Conclusions: The strategy of combining systemic therapy and local control of metastasis with cryoablation had favorable impact on OS in mRCC even after longer follow up. Cryoabalation is a minimally invasive, well tolerated technique, applicable to the majority of mRCC patients. Prospective investigation or case-control studies to evaluate the incorporation of cryoabalation with targeted systemic therapy in advanced RCC is warranted.

1560 Background: Consumption of cruciferous vegetables is associated with decreased risk of prostate cancer (PCa). 3,3’-diindolylmethane (DIM), an in vivo active compound formed after consumption of cruciferous vegetables, down-regulates the AR, and causes its nuclear exclusion which results in in vitro growth arrest and apoptosis of PCa cells. We conducted a biomarker trial evaluating BR-DIM in pre-prostatectomy patients with the primary objective of measuring DIM levels in prostate tissue and in plasma. Methods: Patients with organ-confined PCa who were candidates for surgery were treated with BR-DIM at a dose of 225 mg orally twice daily for a minimum of 14 days. Patients did not receive any androgen deprivation therapy. Patients received their last dose of BR-DIM the day before surgery. Blood samples for DIM levels were collected pre-treatment and just prior to surgery. DIM concentration was measured in the plasma and in prostate tissue specimens using a validated high-performance liquid chromatographic method with tandem mass spectrometric detection. AR was evaluated by immunohistochemistry (IHC). Results: 36 patients were treated at 2 institutions. The 26 evaluable patients had median age of 60 years (range 41 - 76), 14 were Caucasian, and 10 were African American. Reasons for inevaluability included change in surgery date (n=4), inadequate BR-DIM treatment (2), withdrawal from study (2), canceled surgery (1), and other (1). Toxicity was minimal; only two patients with grade 3 headache. Post dosing, DIM was found at a mean trough level of 10.2 ng/ml (range, 0.5 to 24.7) and 12.3 ng/gm of tissue (range, 0.0 to 26.8) in plasma and in prostate tissue, respectively. PSA levels had a slightly downward trend after BR-DIM treatment. In 18 of 20 evaluable PCa specimens, IHC showed nuclear exclusion of AR. Conclusions: BR-DIM was well tolerated, and DIM was detected in both plasma and prostate tissues at ~12 h post dosing. Nuclear exclusion of AR was found in 90% of PCa specimens post BR-DIM dosing, suggesting in vivo inactivation of AR activity. PSA levels were slightly reduced overall. Accrual is ongoing and nearly complete. Additional studies with BR-DIM in PCa are warranted.

415 Background: Pazopanib is an oral tyrosine kinase inhibitor and is currently approved for treatment of patients with advanced or mRCC in the first-line, or second-line setting, after progression on cytokine therapy. No reports of efficacy and safety are available post targeted therapy. We conducted a retrospective analysis of cases receiving pazopanib after failure of other targeted agents. Methods: The study was reviewed and approved by our Institutional Review Board. Review of medical records of advanced renal cancer patients, treated with second-line pazopanib from April 2010 to August 2011 at Karmanos Cancer Center was performed. The Common Terminology Criteria for Adverse Events version 4.0 was used to grade the severity of definitely treatment-related adverse events (AEs). Results: 14 patients are included to date in this ongoing data collection and analysis study. Median age was 58 years, 8 were Caucasian, 4 African American, and 2 other races. By the MSKCC criteria, 1 patient was favorable, 6 were intermediate and 7 poor risk. Pazopanib was utilized as second line in 4 patients, third line in 5, and even later in 5. Prior line targeted therapies included sunitinib (8), temsirolimus (9), everolimus (4), sorafenib (1), bevacizumab (1) and axitinib (1). Nine of 14 patients to date were treated with pazopanib for at least 2 months and were assessed for response by imaging studies. One achieved partial remission; 5 had stable disease, and 3 had progression. The median time to progression after starting pazopanib was 4.0 months. The predominant AEs were grade 1 or 2, which included fatigue (4), liver enzyme elevation (3), diarrhea (2), hypertension (2), skin rash (1), depression (1) and hyponatremia (1). No grade 3 or 4 AEs, or definitely treatment-related deaths were observed. Conclusions: Pazopanib appears to maintain efficacy and tolerability in mRCC even after failure of other targeted therapies. Further investigations of larger patient cohorts in this setting are recommended.

This Phase II study was undertaken to access the activity of recombinant human macrophage colony stimulating factor (M-CSF) in metastatic soft tissue sarcoma based on the observation of a partial response in a patient with leiomyosarcoma during an earlier Phase I trial. Fifteen patients with metastatic soft tissue sarcoma (seven males and eight females) were entered on the trial between October 1990 and March 1991. Seven of these patients had leiomyosarcoma. One mg/M2 of M-CSF was administered by rapid intravenous infusion every 8 h on days 1-5 and 15-19. Treatment cycles were repeated at 35-day intervals. Patients were evaluated initially for response after the first cycle, and then following alternate cycles. One partial response was observed in a patient with metastatic small bowel leiomyosarcoma (response rate 7%, 95% confidence interval, 0-33%). Two additional patients had stable disease for 10-15 months on study. Four patients had clinically significant bleeding from tumor sites during M-CSF therapy. No evidence of toxicity directly attributable to M-CSF was observed in any patient. Mean monocyte counts increased in patients during the first 20 days of treatment (p = 0.013). At this dose and schedule, meaningful activity of M-CSF in previously treated patients with soft tissue sarcoma could not be demonstrated. However, the activity observed in patients with leiomyosarcoma in this trial and in the previous Phase I study are intriguing. Further studies of M-CSF in previously untreated patients with leiomyosarcoma may be warranted.

Breast cancer is a major cause of death and disability among women in the United States. One in nine women will develop breast cancer at some point in their lives. Over the years, treatment modalities have improved, but mortality from breast cancer has changed little since 1930.1 This observation dictates a need for greater efforts toward disease prevention. One important new prevention strategy to reduce the incidence of breast cancer is dietary intervention.

e15185 Background: DIM acts as an anti-androgen which down-regulates the androgen receptor (AR) and prostate specific antigen (PSA). We conducted a biomarker trial to measure the level of DIM and i...

3000 Background: The nuclear enzyme PARP is essential in recognition and repair of DNA damage. In preclinical models, PARP inhibitors work as sensitizing agents for DNA-damaging agents such as Ir. V is an orally bioavailable PARP 1 and 2 inhibitor. This phase I study determined the maximum tolerated dose (MTD), dose-limiting toxicities (DLTs), PK and PD of V together with Ir. Methods: Eligibility included patients with performance status 0-2; ≥ age 18; adequate bone marrow, hepatic and renal function. Cycles were 21 days. Ir was given i.v. 100 mg/m2 over 90 min on Days 1 and 8. Twice daily (BID) oral dosing of V (10-50mg) occurred Days 3-14 (Cycle 1) and Days -1-14 (subsequent cycles) followed by a 6-day rest. PK studies were conducted with both agents alone and in combination. Tumor biopsies and matched peripheral blood mononuclear cells (PBMCs) were collected for PD evaluation of PAR, γ-H2AX, PARP1, TOPO1, and ERCC1 after Ir alone and after the combination. Results: 32 pts were treated (2 lung, 14 breas...

TPS185 Background: Docetaxel based therapy in metastatic castrate resistant prostate cancer (mCRPC) provides an improvement in progression free survival and overall survival. However cures are extr...

33 Background: To this day, no therapy has proven efficacy in the clinical management of PSA relapse non-metastatic CRPC. Investigating novel therapeutic agents that will induce a PSA response, and delay progression to overt metastases is of paramount importance. Methods: A single arm multi-center phase II clinical trial is being conducted. Inclusion criteria are PSA only progression despite androgen deprivation therapy, defined as three rising PSA levels with a measurement interval of at least 2 weeks, and minimum PSA level ≥ 1 ng/mL. Patients with radiologic evidence of bone metastases were excluded. Bevacizumab single agent was administered at 10 mg/kg every 14 days. PSA was evaluated at least every 6 weeks and bevacizumab was continued until PSA progression or new metastasis. Results: 15 patients (see Table ) have been enrolled on this ongoing trial. Toxicities observed were grade 4 pulmonary embolism, grade 3 hypertension, and grade 3 proteinuria in one patient each, but no treatment related mortality. Median treatment duration was 12.6 weeks (range 11–33.1 wks). 12 patients are evaluable for PSA response. Four patients had PSA decline of 14, 20.9, 24, and 45% respectively. Median PSA doubling time during bevacizumab increased to 5.93 months (range 0.81–27.09 months) from pretherapy median of 4.08 months, a 45% improvement. Five patients have progressed to develop bone metastases at median of 7.7 months (range 2.8–18.2) after protocol registration. Conclusions: Bevacizumab was well tolerated, and impacted PSA kinetics in non-metastatic CRPC. Further investigations are warranted. Supported in part by Genentech Inc. (clinicaltrials.gov NCT00478413 ). [Table: see text] [Table: see text]

152 Background: Satraplatin is an oral platinum that has demonstrated efficacy and tolerability in metastatic CRPC. Bevacizumab has revealed safety and efficacy in advanced prostate cancer, and synergy was noted between platinum based chemotherapy and B. Methods: Primary endpoint was time to progression (TTP). Latter wasdefined per RECIST 1.0 or onset of a skeletal event, or > 2 new areas of bone metastases. DP metastatic CRPC patients were eligible. S 80mg/m2 orally on days 1-5, P 5 mg twice daily, and B 10mg/kg on day 1, and 15mg/kg on day 15 were administered in 35 day cycles. Results: 31 patients enrolled (13 African American and 18 Caucasian) to complete accrual. Median age was 67 years (range 50-85 years) and 21 patients (68%) were > 65 years of age. Median pretherapy PSA was 180.7 ng/ml (range 4.7-1,433 ng/ml). 21 (68%) had bone pain, Gleason score was > 8 in 20 (65%) patients. Pretherapy 12 patients had measurable disease progression, 17 (55%) had bone scan progression, and 8 had PSA only progression. 176 cycles have been administered; median 4 cycles (range 0-12 cycles). Grade 4 toxicities noted were, pulmonary embolism in 2 patients and thrombocytopenia in 1 patient. Grade 3 toxicities observed were neutropenia and hypertension in 3, anemia in 7 and , thrombocytopenia and diarrhea in 2 patients each. No treatment related deaths. 29 patients are response evaluable to date; 10 (34%) had a ≥30% PSA decline and 3 (10%) had a > 90% PSA decline. Of 12 patients with MD, 2 had a response and 7 had stable disease. Median TTP was 7.4 months (90% CI 4.8-12.8 months) and median survival was 11.2 months (90% CI 9.1-18.3 months). 47% of patients were alive at 12 months. Genotype characterization for excision repair cross-complementation group 1 (ERCC1) polymorphism was performed in 17 patients with 9 having homozygous (CC), 3 with heterozygous, (CT) and 2 patients with absence of ERCC expression respectively. Conclusions: The combination was tolerable and revealed promising efficacy in metastatic CRPC. ERCC1 testing will be correlated with outcome endpoints. Supported in part by Genentech Inc and GPC Biotech. [Table: see text]

Long-term (1 y) effects of dietary fat intake on lipoprotein metabolism were determined in 72 healthy women receiving either a 15%-fat diet (n = 34) or usual diet (n = 38). Every three months food records, weight, waist-hip ratio (W:H), percent body fat, fasting plasma triglyceride, cholesterol (C), high-density-lipoprotein cholesterol (HDL-C), HDL2-C, and HDL3-C; apolipoprotein B and A-I, and postheparin lipoprotein lipase (LPL) and hepatic triglyceride lipase activities were determined. In one year, the low-fat-diet (LFD) group had 17% and the non-intervention-diet group had 36% dietary fat. The LFD group showed decreases in cholesterol: 7% TC, 13% low-density lipoprotein (LDL), and 8% HDL. Apolipoprotein A-I, decreased early. Apolipoprotein B did not change. Plasma triglyceride correlated with weight. Percent body fat and W:H correlated with the total and LDL-C. Changes in HDL-C and/or HDL2-C and LPL correlated directly with the changes in dietary fat and inversely with dietary carbohydrate. Changes in total-C or LDL-C correlated with the changes in weight and W:H, but not with the changes in nutrient intake.

e14120 Background:Information on the prognostic and predictive role of FDG-PET in the management of patients (pts) with mCRC is limited. The growing complexity of current therapies and the increasi...

Increased fat and caloric content of the diet has been associated with increased mammary tumor incidence. The dietary modulation of cellular redox state may be one mechanism behind this association. We have examined the effects of changes in dietary fat and caloric intake on the levels of 5-hydroxymethyluracil in DNA from rat liver and mammary gland. Female Fischer 344 rats, 40 days old, were maintained on 3% (low-fat), 5% (control), or 20% (high-fat) corn oil diets for 2 weeks. A fourth group of rats had the same daily fat intake as the control group, but total caloric intake was restricted by 40%. As a measure of oxidative DNA damage, 5-hydroxymethyluracil levels were measured in the DNA extracted from liver and mammary gland by gas chromatography-mass spectrometry. 5-Hydroxymethyluracil levels in the liver DNA of the low-fat, high-fat, and calorie-restricted groups were decreased relative to that of control, but the only significant decrease was in the calorie-restricted group (p less than 0.01). In the mammary gland DNA, statistically significant decreases in damage were found in each group relative to control (p less than 0.05). The relationship between fat in the diet and oxidative stress is thus complex. These results show that changes in dietary intake of both fat and calories can modulate oxidative DNA damage levels, and the effect of diet was more clearly evident in the DNA from mammary gland than in DNA from liver.

e20626 Background: It is important to increase our understanding of recruitment barriers in the elderly population to improve their participation in clinical trials. The primary objective of this study was to describe the demographics, treatment, toxicity, and overall survival (OS) of all patients ≥ 65 years who presented to the Phase I clinical trials service at KCI between 1995–2005. Methods: Patient demographics, co-morbidities, tumor type, reason not enrolled on a Phase 1 clinical trial, toxicities, and OS were obtained by retrospective chart reviews. The patients were divided into 3 mutually exclusive groups: patients considered but not enrolled (PC), enrolled but not treated (PE), and treated (PT). OS was compared across these three groups. Results: 216 patients met the study criteria. Of these, there were 114 PC, 7 PE, and 95 PT. Lack of enrollment was due to ineligibility (30%), loss to follow-up (12%), refusal (8%), or no trial available (2%). Median age was 71 years, 63% were males and 87% were Caucasian. 59% had performance status (PS) 1; 16% had PS 2. 66% of patients had a history of cardiovascular disease but renal, liver, hematological diseases were found in less than 7% of patients at baseline. Colorectal (27%), lung (15%) and prostate (8%) were the 3 most common cancers. 54% received a cytotoxic agent, 62% a biologic agent, and 17% combination therapy. The 3 most common Grade 3–4 toxicities were: 15% leucopenia, 11% electrolyte abnormalities, and 4% anemia. The median OS for PC, PE, and PT was 3.9 mos, 2.2 mos, and 8.4 mos, respectively (p < 0.001 between any pair). However, PS 0–1 patients comprised 47%, 75%, and 73% of those 3 patient subgroups, respectively. Conclusions: More than half of the elderly patients initially considered for Phase 1 clinical trials at KCI were not enrolled due to ineligibility, refusal, or lost to follow up. Phase I chemotherapy appears feasible in patients age ≥ 65 years old. OS was significantly longer for patients treated on Phase I trials compared to those who were not, although that may be confounded by PS differences. More detailed OS comparisons will be presented. No significant financial relationships to disclose.

e19007 Background: There are only limited data regarding the use of anti-VEGF therapy in recurrent NSCLC and no data in NSCLC pts previously treated with bevacizumab. We are currently conducting a phase II trial evaluating cediranib, an oral inhibitor of VEGFR 1,2 and 3, and pemetrexed in recurrent NSCLC pts who may or may not have previously received bevacizumab. Methods: Pts with progressive and measurable NSCLC, 1 or 2 prior regimens, PS0–2, all histologic sub-types, BP ≤ 140/90, treated brain metastases are eligible. Pts on anti-coagulants are allowed. Pts with hemorrhage within 4 weeks are excluded. Pts start on cediranib 30mg daily followed 7 days later by pemetrexed at 500 mg/m2 every 21 days and cediranib daily. The study consists of two cohorts- cohort A (no prior bevacizumab) and cohort B (prior bevacizumab). Planned accrual is 37 pts each cohort. Consenting pts will undergo FLT PET scans and blood draw for circulating tumor cells before therapy, 1 week after cediranib, and after 1 cycle of the combination. Results: 33 pts have started therapy, (Cohort A- 20, Cohort B- 13), median age- 60, males- 56%, ever smokers- 88%, adenocarcinoma- 64%, squamous- 12%, brain mets- 27%, 1 prior regimen- 52%, PS0–1- 88%. Median cycles- 4 (range- 0–15). Grade 3/4 toxicities- neutropenia- 7pts, febrile neutropenia- 1pt, fatigue-7pts, diarrhea- 3pts, hypertension- 1pt, anorexia- 2pts, cardiac ischemia- 1pt, bronchopleural fistula- 1pt, esophagitis- 1pt. No major hemorrhage. Of the 17 pts who received cediranib for ≥ 4 cycles, 71% required dose reduction and of the 18 pts who received pemetrexed for ≥ 4 cycles, 22% required dose reduction. 31 pts (Cohort A- 19, Cohort B- 12) are response evaluable. Confirmed response rate is 16%(90% CI- 0.08–0.30) (Cohort A- 10%, Cohort B- 25%) and disease control rate (response+stable disease) is 71% (90% CI-0.56–0.82) (Cohort A- 74%, Cohort B- 67%). 8 of 9 pts who had FLT PET scans had a 20% or greater decline in standard uptake value after 1 week of cediranib alone. Conclusions: Cediranib and pemetrexed combination is tolerable. Efficacy has been observed with the combination in recurrent NSCLC pts, including those previously treated with bevacizumab. Accrual to this trial is ongoing. [Table: see text]

5565 Background: Both D and G have demonstrated single-agent activity in OC and PC. The purpose of this study was to prospectively evaluate the efficacy and toxicity for patients (pts) with relapsed platinum-sensitive (Plat-S) and platinum-resistant (Plat-R) OC or PC treated with combination weekly D and G. Methods: Eligibility criteria included women recurrent or refractory to first line platinum based therapy, a performance status of 2 or better, and adequate renal and hepatic function. Only one prior regimen (including maintenance therapy) was allowed. D (40 mg/m2) and G (800 mg/m2) were administered on days 1 and 8 of a 21 day cycle until progression. Best response was defined by RECIST criteria. Granulocyte stimulating factor prophylaxis was not allowed. There were 2 separate 2-stage designs used, one for each stratum: Plat-R and Plat-S, separately for a total planned sample of 62 pts. The primary endpoint was overall response rate (ORR). This study was approved by each center's institutional review board. Results: Thirty pts were enrolled prior to terminating accrual due to lagging enrollment. The median age was 57.5 years (range 41–80). One pt was diagnosed with PC; the remaining 29 had OC. Twenty-seven (90%) pts were response evaluable. The ORR was 59% (90% confidence interval: 0.44 - 0.73). Six pts (22%) achieved a complete response (CR), 10 pts (37%) achieved a partial response (PR) and one pt (4%) had stable disease. Progressive disease was noted in 10 pts (37%). Plat-S pts (n = 16) had a 69% ORR (CR = 31%, PR = 38%) and Plat-R pts (n = 11) had an ORR of 45% (CR = 9%, PR = 36%). Median overall survival was 12.7 months (Plat-S = 14.2 mos. and Plat-R = 6.7 mos.). Toxicity data were evaluable for 29 (97%) pts. Twelve pts (41%) experienced grade 3 or 4 neutropenia; two of these pts also had documented infections. Three pts had grade 3 anemia and 6 (21%) had grade 3 thrombocytopenia. Dose reductions were uncommon (n = 2, 7%). Twelve pts (41%) were hospitalized at least once during treatment. Conclusions: Weekly D and G combination therapy demonstrates significant activity and moderate toxicity in both Plat-S and Plat-R disease. This non-platinum combination deserves further evaluation and may be considered in pts with Plat-S and Plat-R OC and PC. No significant financial relationships to disclose.

5108 Background: S, T (temsirolimus) and E (everolimus) are primarily metabolized in the liver, while the metabolism of B is unclear. There are limited data on the clinical toxicity profile and efficacy of these agents in pts with RI. Methods: The primary objective was to assess the safety and efficacy of S, B, T and E in pts with RI. Medical records of pts with mRCC at Wayne State University, treated on S, B, T or E were reviewed. Pts with a calculated creatinine clearance (CrCl) of ≤ 60ml/min [chronic kidney disease stage 3 or higher per K/DOQI guidelines by the National Kidney Foundation] were deemed to have RI. Data on safety and efficacy of the therapy were collected and analyzed with respect to renal function. Results: 19 of 51 (37%) pts had RI. Pts with RI had a higher median rise in blood pressure (BP) with S and B than pts with normal renal function. Patients with RI had an increased incidence of rash and higher dose interruption rates with m-TOR inhibitors. No major differences in toxicities including cardiac, thyroid, renal, lipid profile abnormalities or hyperglycemia were observed. Similar efficacy was seen in all groups. Conclusions: More than a third of pts with mRCC receiving targeted therapy have RI, hence highlighting the importance of evaluating tolerability of therapies in pts with RI. Therapy with S, B and T/E is well tolerated and efficacy appears to be maintained. Closer monitoring for hypertension is needed in pts receiving S and B. [Table: see text] No significant financial relationships to disclose.

e15037 Background: Information on the prognostic and predictive role of FDG-PET in the management of patients (pts) with mCRC is limited. The growing complexity of current therapies and the increasing number of agents to be tested in this disease warrants better understanding of the role of FDG-PET in earlier treatment decisions. Methods: Consecutive pts with 2 or more serial FDG-PET scans at baseline and during the treatment course were studied. Tumor standardized uptake value (SUV) and its percentage change (%ΔSUV) were each studied for their potential association with time to progression (TTP) via univariate Cox models to estimate the hazard ratio (HR) for progression. Results: 27 pts (median age 58.2 yrs) with mCRC were studied. 85% of pts were treated in the first line setting. 44% had received prior adjuvant therapy. 63%, 26% and 11% received oxaliplatin based, irinotecan based and fluropyrimidine only regimens, respectively. 85% received concurrent bevacizumab. Median pretreatment SUV was 9.0 (range 1.7 - 46.0); Median post treatment SUV was 3.4 (0–13.5); median %ΔSUV was -77.2 (range -10% to -100%). Mean interval between scans was 4.1 months. Ten (37%) patients had no tumor uptake on post treatment scans. 56% and 37% of pts had partial response and stable disease (RECIST criteria), respectively. Median TTP was 13.0 months (90% CI: 10.9 - 16.3 mos), with a median follow-up time for progression of 7.8 months. The HRs for baseline SUV and %ΔSUV were 0.972 (90% CI: 0.901 - 1.048, p=0.534) and 1.018 (90% CI: 1.003 - 1.033, p=0.049), respectively. The median TTP of patients whose post-treatment SUV reached zero was 13.8 months vs. 10.9 months (p=0.17) for pts whose post-treatment SUV did not reach zero. Conclusions: Systemic therapy significantly decreased the SUV on follow up PET scans in pts treated for mCRC. However, no significant association was seen between either baseline SUV or %ΔSUV and TTP. There may be a very weak statistical association of decreasing SUV with decreasing risk of progression. Further work is needed to optimize and standardize evaluation of tumor response in mCRC patients with FDG-PET. No significant financial relationships to disclose.

e15586 Background: Human immunodeficiency virus (HIV) infected patients (pts) are at increased risk for squamous cell carcinoma of the anal canal (SCCAC) and the incidence of SCCAC has increased in the era of HAART (highly active antiretroviral therapy). The outcome of SCCAC in HIV (+) pts has not been evaluated in prospective trials and the published literature is limited to small retrospective case series. The aim of this study is to describe the outcome, tolerability, and overall survival (OS) in pts with and without HIV infection treated at Karmanos Cancer Institute, at Wayne State University from 1991 to 2007. Methods: We performed a retrospective chart review. We collected data regarding HIV status, demographics (age, gender, race), stage at diagnosis, treatment, response to treatment, toxicity, and survival. Results: Fifty pts with SCCAC were identified, of whom 18 were HIV (+) and 32 were HIV (−), 26% Caucasians, 68% African American, 56% males and 44% females. HIV (+) pts had significantly better stage (p = 0.011) and less frequent reduced chemotherapy dose (p = 0.001). There were no significant differences by HIV status in type of chemotherapy received, frequency of reduced radiotherapy dosage, use of diverting colostomy, or frequency of relapse. Diverting colostomy was required in 29% of HIV (+) pts and 18% of HIV (−) pts. The major toxicities observed in HIV (+) and (−) pts were diarrhea (36% vs. 64%), neutropenia (27% vs. 21%), and skin toxicity secondary to radiotherapy (XRT: 82% vs. 100%; p = 0.034). Median (OS) was 62.6 months for HIV (+) pts and 71.8 months for HIV (−) pts (p = 0.787). Conclusions: HIV (+) pts had better stage, received standard chemotherapy dose more often, and had more frequent XRT dermatitis than HIV (−) pts. Otherwise, there was no major difference in treatment toxicities. A higher proportion of HIV (+) pts required diverting colostomy. Survival is somewhat shorter among HIV (+) patients. No significant financial relationships to disclose.

4563 Background: VEGF (vascular endothelial growth factor) has a central role in angiogenesis, tumor growth and metastasis of gastric cancer. Bevacizumab, an anti-VEGF monoclonal antibody, has demonstrated anti-tumor activity in multiple diseases. This phase II study was undertaken to determine the effects of adding bevacizumab to a regimen of docetaxel and oxaliplatin. Methods: The primary endpoint was time to progression (TTP) in patients with locally advanced or metastatic adenocarcinoma of the gastric or gastroesophageal junction treated with docetaxel, oxaliplatin and bevacizumab. Previously untreated patients with a performance status (PS) of 0–1 were eligible for this study. Patients received bevacizumab 7.5 mg/kg, docetaxel 70 mg/m2 and oxaliplatin 75 mg/m2 administered on day 1 of a 21 day cycle. Results: A total of 23 patients (median age 57, males 70%, gastric 52%) were enrolled on the study. Median PS was 1. The median number of cycles was 5. Ten patients are still receiving treatment on study. Partial responses were documented in 10 (59%) patients and stable disease in 7 (41%). No treatment related deaths were observed. The most commonly reported grade 3–4 toxicities were neutropenia (13%), leukopenia (4%), fever (4%), acute neuropathy (4%), and hypertension (4%). Gastrointestinal (GI) perforation occurred in 3 patients. Perforation was not found at the tumor site in the patient who required surgery. The site of perforation could not be ascertained in the second patient who was managed medically. Both patients had had no prior surgical resection of the primary tumor. The third perforation presented as a tracho-bronchial fistula. The patient had previously undergone surgical resection of his primary tumor after receiving chemoradiotherapy to the thoracic area. Conclusions: The regimen of docetaxel, oxaliplatin and bevacizumab appears to be very active. The development of GI perforations in 3 patients is of concern. At this time, bevacizumab should not be used in gastric or gastroesophageal junction cancers outside of a clinical trial until its safety is well established. [Table: see text]