Search

Abstract 3286 Acute myeloid leukemia (AML) cells express the cell surface antigen CD33 that is a down-regulator of cell growth when ligated by a monoclonal antibody in a Syk-dependent manner. The response of AML cells to gemtuzumab ozogamicin (GO) also depends on Syk and SHP-1 expression (Leukemia 20:2093, 2006). The hypomethylating agent 5-azacytidine (5-aza) induced re-expression of Syk in some cases, therefore increasing the sensitivity of originally Syk-negative, non-responsive cells to CD33 ligation to levels of Syk-positive cells. We initiated a phase 1/2 clinical trial examining if treatment with 5-aza prior to GO is safe, efficacious, and whether in vivo responses to GO correlated with Syk expression and induction by 5-aza. Here we update the interim results of this trial (NCI registration number NCT00766116). In Phase I, 14 patients (9 males, 5 females), age range: 39–82 years [median: 66]) were treated with 75mg/m2 5-aza daily and GO in a dose-escalation manner, 4 cohorts total. The first cohort (n=3) received 5-aza for 2 days followed by GO at 3 mg/m2 on days 3 and 17; the second cohort (n=3) received 5-aza for 2 days followed by GO at 6 mg/m2 on days 3 and 17; the third cohort (n=4) received 5-aza for 4 days followed by GO at 6 mg/m2 on days 5 and 19; and the fourth cohort (n=4) at 5-aza for 6 days followed by GO at 6 mg/m2 on days 7 and 21. There were no responses in the first 2 cohorts. One patient in cohort 3 achieved CR, and 2 in cohort 4 achieved CR and CRp. Adverse events (≥ Grade 3) included febrile neutropenia 36%, infection 14%, pancytopenia 7%, dyspnea 7%, and retinopathy 7%. Average length on study (n=14) was 45 days with a mortality rate of 14% (unrelated to treatment). No dose-limiting toxicities were encountered in phase I, therefore the MTD is the dose in cohort 4. The overall response rate in evaluable patients in phase I (n=11) is 27%. Average time to ANC recovery (n=6): 30 days (range 15–42, median 33 days). In Phase II, 10 patients (5 males, 5 females), age range: 29–64 years (median 60) were treated at the MTD: 5-aza for 6 days and GO at 6 mg/m2 on days 7 and 21. 8 patients were in 1st relapse, 1 in 2nd and 1 in 3rd. There were 3 responders (2 CR, 1 CRp) in this phase, all in 1st relapse at baseline. Adverse events (≥ Grade 3) include febrile neutropenia 50%, infection 20%, increased LFTs 10%, thrombocytopenia 10%, dyspnea 10%, wheezing 10%, mucositis 10%, cough 10%, and hypoalbuminemia 10%. The average length on study (n=10) was 40 days with a mortality rate of 10% (not related to study treatment). Average time to ANC recovery in phase II (n=2): 15 days (range 12–17, median 15) with an overall response rate in evaluable patients (n=7) of 43%. The ORR for phase I/II (n=18) is 33%. 21 of the 24 patient sample pairs have been analyzed for Syk and SHP-1 expression (one patient did not have a baseline sample). Prior to therapy, Syk was expressed in 16 of 20 cases. After 5-aza treatment, Syk was re-expressed in all 4 negative cases, and increased over baseline in one case that was previously Syk +. SHP-1 was positive in 17 of the 20 cases and was re-expressed in all 3 negative cases. Leukemia cells from patients who achieved CR were Syk+ in 3 of 5 cases (the 6th hasn't been analyzed). Syk was re-expressed in the two negative cases after 5-aza. SHP-1 was expressed in 4 of 5 cases at baseline, and re-expressed in the one negative case after 5-aza. In vitro we analyzed inhibition of proliferation (for patients 1–6) or colony formation (for patients 7–24) induced by 5-aza and GO. 5-aza alone allowed 62.3+/−3.5 survival of leukemia cells and GO alone allowed survival of 59.5+/−1.7 leukemia cells. However, exposure to both agents resulted in a survival rate of 24.8+/−1.6 (P Disclosures: Ball: Celgene: Equity Ownership, Research Funding. Off Label Use: Will discuss use of 5-azacytidine (Vidaza) for treatment of relapsed AML in combination with Mylotarg (on label, but only as monotherapy). Medeiros:Celgene: Research Funding, Speakers Bureau; Novartis: Research Funding, Speakers Bureau; Merck: Research Funding; Genentech: Research Funding; Alexion: Speakers Bureau.

9162 Background: Cancer-related fatigue is a widely reported distressing symptom. We examined biochemical correlates of fatigue in patients with and without cancer at Moores UCSD Comprehensive Cancer Center (MCCC). Methods: All patients at first visit to MCCC complete a 39-question screen assessing problems including fatigue, sleep and mood disturbance on a 5-point scale (0=not a problem, 1=mild, 2=moderate, 3=severe, 4=very severe). We measured cortisol, IL-1B, IL-6, IL-8, neopterin, TNF- a, testosterone and tryptophan as potential biomarkers of fatigue in blood samples collected the same day for the MCCC biorepository. ANOVA was used to assess differences in mean concentration among 0, 1 and 2+ fatigue levels, overall and for cancer and non-cancer patients. Results: Overall, 118 (61%) of 195 [133 cancer and 62 non-cancer] patients reported problems with fatigue. There was no significant difference in the proportion of cancer (n=81 ; 61%) and non-cancer patients (n=37 ; 60%) reporting fatigue. Tryptophan...

Consider the problem of estimating the value of a functional $\Lambda(f)$ for $f$ an unknown density or regression function. The straightforward plug-in estimator $\Lambda(\hat f)$ with $\hat f$ a particular estimate of $f$ achieves the optimal rate of convergence in the sense of Stone over bounded subsets of a Sobolev space for a broad class of linear and nonlinear functionals. For many functionals the rate calculation depends on a Frechet-like derivative of the functional, which may be obtained using elementary calculus. For some classes of functionals, $\hat f$ is undersmoothed relative to what would be used to estimate $f$ optimally. Examples for which a plug-in estimator is optimal include $L^q$ norms of regression or density functions and their derivatives and the expected integrated squared bias. When interested in computing estimates over classes of functions which satisfy certain restrictions, such as strict positivity or boundary conditions, the plug-in estimator may or may not be optimal, depending on the functional and the function class. The functional calculus establishes conditions under which the plug-in estimator remains optimal, and sometimes suggests an appropriate modification when it does not.

Abstract 2049 Poster Board II-26 Acute myeloid leukemia (AML) cells express the cell surface antigen CD33 that, upon ligation with a monoclonal antibody (mAb), is a downregulator of cell growth in a Syk-dependent manner. An anti-CD33 mAb coupled to calechiamycin, gemtuzumab ozogamicin (GO), is used for the treatment of AML. We demonstrated that the response of AML cells to GO treatment also depends on Syk and SHP-1 expression (Leukemia 20:2093, 2006). Syk upregulation by the demethylating agent 5-azacytidine (5-aza) induced re-expression of Syk in some cases, resulting in enhanced GO and anti-CD33-mediated inhibition of leukemia cell growth. Thus, the cytotoxicity of both GO and anti-CD33 in primary AML samples was associated with Syk expression. 5-Aza restored Syk and increased the sensitivity of originally Syk-negative, non-responsive cells to CD33 ligation to levels of Syk-positive cells. We designed a clinical trial examining if treatment with 5-aza prior to GO is safe, efficacious, and whether in vivo responses to GO correlated with Syk expression and induction by 5-aza. Here we report the interim results of this trial. In Phase I, 14 patients (9 males [7 Caucasian, 1 Asian, 1 Hispanic], 5 females [2 Caucasian, 3 Asian], age range: 39-82 years [median: 66]) were treated with 75mg/m2 5-aza and GO in a dose-escalation manner: the first cohort (n=3) received 5-aza for 2 days followed by GO at 3 mg/m2 on days 3 and 17; the second cohort (n=3) received 5-aza for 2 days followed by GO at 6 mg/m2 on days 3 and 17; the third cohort (n=4) received 5-aza for 4 days followed by GO at 6 mg/m2 on days 5 and 19; and the fourth cohort (n=4) at 5-aza for 6 days followed by GO at 6 mg/m2 on days 7 and 21. No dose-limiting toxicities have been encountered. There were no responses in the first two cohorts (5 patients in 1st rel, 1 in 2nd rel at start of treatment); one patient in cohort 3 achieved a CR (in 1st relapse), two in cohort 4 have achieved CR (1 in 1st rel and 1 in 2nd rel). We are in Phase II of the study: Cohort 4 dose 5-aza for 6 days followed by GO at 6 mg/m2 on days 7 and 21. Patients remained on study until ANC recovered to 1000/mm3 for 3 consecutive days and were assessed for response to treatment. Average time to ANC recovery (n=6) 30 days (range 15-42) (median 33 days). Average length of time on study: 46 days. Prior to therapy, Syk was expressed in 10 of the 13 cases (baseline data not available on one case). After 5-aza treatment, Syk was re-expressed in each of the 3 negative cases, and increased over baseline in one case that was previously Syk +. SHP-1 was positive in 10 of the 13 cases and was re-expressed in each of the 3 negative cases. The leukemia cells from the patients who achieved CR were Syk+ in 2 of 3 cases. Syk was re-expressed in the one negative case after 5-aza. SHP-1 was expressed in 2 of 3 cases at baseline, and re-expressed in the one negative case after 5-aza. In vitro we analyzed inhibition of proliferation (for patients 1-6) or colony formation (for patients 7-14) induced by 5-aza and GO treatment. In all cases 5-azacytidine alone mediated about 50%, and GO alone about 40% cytotoxicity. Together they killed about 80% of leukemia cells. We also compared pre- and post 5-aza samples from the same patients: in all cases 5-aza treatment increased the GO-mediated cytotoxicity. These initial data show that in vivo exposure to 5-aza can induce the expression of two biomarkers involved in the intracellular response to GO. Correlations with response will be made in the Phase II portion of the study now underway. Disclosures: Ball: Celgene: Research Funding. Off Label Use: 5-azacytidine used in treatment of AML.

9608 Background: Cancer-related fatigue is a widely reported distressing symptom. We examined patterns and clinical correlates of fatigue among a series of patients with and without cancer at Moores UCSD Comprehensive Cancer Center (MCCC). Methods: All patients who come for initial consultation to MCCC complete a 39-question screen assessing problems including fatigue, sleep, and mood disturbance on a 5-point scale (0 = not a problem, 1 = mild, 2 = moderate, 3 = severe, 4 = very severe). We report prevalence of fatigue from 1,880 consecutive patients; the charts of 515 of these were reviewed for further data. Results: Overall 1,137 of 1,880 (61%) patients reported problems with fatigue. Of the 515, there were 321 cancer and 194 non-cancer patients. 40 patients not responding to the fatigue question were excluded. There was no significant difference in the proportion of cancer (189 of 296; 64%) and non-cancer patients (104 of 179; 58%) reporting fatigue. Prevalence was similar between genders. 126 of 208 (61%) chemotherapy-naïve cancer patients reported fatigue compared to 63 of 88 (72%) patients who received any chemotherapy (p=0.07). In comparison to the chemotherapy-naïve, 37 of 48 (77%) with ongoing or recent (< 6 mo) chemotherapy reported fatigue as a problem (p=0.03); but among patients who underwent chemotherapy > 6 mo prior to the questionnaire, fatigue was not significantly different (26 of 40; 65%). 51 of 189 (27%) cancer patients married or living with a partner reported moderate to severe fatigue compared to 39 of 104 (38%) with no partner (p=0.06). Among cancer patients without sleep or mood problems, 19 of 71 (27%) reported fatigue as a problem compared to 166 of 221 (75%) with sleep and/or mood problems (p < 0.001). Conclusions: Fatigue is prevalent among cancer and non-cancer patients. Although more cancer patients with ongoing/recent chemotherapy reported fatigue compared to the chemotherapy-naïve, after 6 mo there was no significant difference, suggesting that fatigue may only be partly explained by chemotherapy. Social support from a spouse/partner may also help reduce problems with fatigue. Sleep and mood disturbance are associated with fatigue, but many cancer patients reported fatigue in its absence, consistent with previous data and suggestive of other etiologies. No significant financial relationships to disclose.

It has been observed that to a smoothing spline operator there corresponds an equivalent kernel operator; these two operators have been compared in a variety of norms [Cox (1984), Silverman (1984)]. In this paper, we refine the existing bounds for the particular case of the spline estimator considered in Rice and Rosenblatt (1983) and its corresponding equivalent kernel estimator. We obtain detailed asymptotic expressions for the bias and covariance functions of the two estimates and provide rates of convergence. Direct comparison then shows that the two estimates are similar. They may differ somewhat in their higher order boundary behavior.

Despite the introduction of modern therapeutics, overall survival (OS) in metastatic breast cancer (MBC) has not changed in 20 ys. In the 1990s a major effort was made to improve the situation through the use of high dose chemotherapy with autologous stem cell transplantation (HD-ASCT). Although OS was not affected in randomized trials, which led to near abandonment of this therapy, some studies suggested that progression-free survival (PFS) can be improved with HD-ASCT. To identify distinct transplant-, disease- and patient-related characteristics predictive of survival in patients with MBC who received HD-ASCT, we reviewed records of all patients in the bone marrow transplant registry at UCSD treated with HD-ASCT for MBC between 1989 and 2000. METHODS: Age, race, stage at diagnosis, histology, estrogen receptor (ER) and menopausal status, body mass index (BMI) in kg/m2, time to transplant and death, site of metastasis, disease status prior to and after transplant, and days in hospital were extracted from medical records. Brookmeyer & Crowley’s 95% confidence intervals were used for median survival times, Cox models for predictors of a time-to-event variable and Schoenfeld tests for proportional hazard assumptions, respectively. RESULTS: Of 96 patients with MBC 21, 43, 23, 8 and 1 had stage I, II, III, IV or unknown disease at diagnosis, respectively. Histologies were: infiltrating ductal, lobular, inflammatory or unknown in 79, 10, 6 and 1 of the cases, respectively. ER status was positive in 59.4% of patients and 29.2% of patients were post-menopausal. 84.4%, 4.2% and 11.5 % of patients were of caucasian, black or other ethnicity, respectively. Mean ages at diagnosis and at transplant were 43.7 ys (SD 8.6) and 47.4 ys (SD 8.7), respectively, with a median time to transplant of 29 months (range 3.8–180.8). When progression of disease (PD) was diagnosed, primary sites were visceral (39.6%), bone (29.2%), local (16.7%) or nodal (14.6%). Mean BMI at transplant was 26 (SD 5.9); 24% of patients were obese (BMI ≥ 30). Mean length of hospitalization after transplant was 17 days (SD 9.5 days); 33% of patients were hospitalized ≥ 18 days. Pre-transplant, 30.2% of patients were in complete remission; after HD-ASCT, this percentage increased to 41.7%. Median PFS and OS were 3.9 ys (CI 2.9–5.4) and 5.6 ys (CI 4.1–7.4) after initial diagnosis and 0.6 ys (CI 0.5–0.8) and 1.7 ys (CI 1.36–2.07) after transplant, respectively. As opposed to ER+ patients the rate of death events in ER− patients slowed down substantially after 5 ys from diagnosis. Although not statistically significant, at 12 ys, survival after HD-ASCT was 18.5% in ER− patients and 2.6% in ER+ patients. Stratified by ER status, stage at diagnosis was an independent predictor of length of PFS and OS. At diagnosis, stage I patients were at lowest risk of death when compared to stage II–IV patients with HRs of 2.7 (II vs I CI 1.4–5.2), 4.6 (III vs I CI 2.1–10) and 17 (IV vs I CI 6.1–47.8) disease; lower risk of death persisted for patients with initial stage I after PD was diagnosed with HRs of 2.4 (II vs I CI 1.3–4.6), 2.8 (III vs I CI 1.3–5.7) and 3.9 (IV vs I CI 1.6–9.9). Death risks were increased with infiltrating lobular carcinoma when compared with infiltrating ductal carcinoma (HR 2.5; CI 1.1–5.38) or when BMI was ≥ 30 (HR 3.1; CI 1.8–5.4); PFS in patients with a BMI ≥ 30 was significantly shorter after PD was diagnosed (death HR 3.1 [CI 1.8–5.5] when compared to BMI < 30). Visceral when compared to bone metastasis was a negative predictor of OS (HR 2.3; CI 1.3–4.1). Hospitalization ≥ 18 days after HD-ASCT was the strongest predictor of time to death after transplant (HR 2.2; CI 1.4–3.6). DISCUSSION: The survival rate at 12 ys after HD-ASCT was 8.2% for patients with MBC. Survival was higher for ER- compared to ER+ patients. Negative prognostic factors for OS and PFS were higher stages of disease at initial diagnosis, but also once PD was diagnosed, infiltrating lobular histology and obesity. Since all transplants were carried out prior to routine assessment of Her-2 neu receptor status, this information was not part of our analysis. At PD, visceral metastasis translated into poorer OS and after HD-ASCT, length of hospitalization became the most important outcome predictor. Although HD-ASCT in MBC has lost favor because of no improvement in global OS, our analysis may provide a rationale for selection criteria to determine which patients could benefit from future trials of HD-ASCT.

Purpose Surgical resection remains the treatment of choice for patients with stage I non—small-cell lung cancer (NSCLC). However, there is high likelihood of medical comorbidity in this patient population, requiring management by nonsurgical approaches. We report our experience using conventional and hypofractionated radiation therapy schedules with conformal approach. Patients and Methods Between 1991 and 2006, 108 patients with medically or otherwise inoperable stage T1/T2 N0 NSCLC were treated with curative radiation therapy alone at our institution. Patient characteristics were as follows: median age, 73 years (range, 37–86 years); male, 88/108 (81.5%); stage T2, 46/108 (42.6%), and histology/cytology, 91/108 (84.3%). Patients received a median total dose of 6500 cGy using median daily dose fractions of 250 cGy. The majority of patients were treated using hypofractionated schedules: daily dose fractions > 200 cGy, 79/108 (73.1%). The following outcomes were analyzed: local failure-free survival (LFFS; time to local failure or death from any cause), time to local or distal failure as first event, and overall survival (OS). Local failure was defined as an increase in size on imaging studies. Toxicities were evaluated using Common Terminology Criteria for Adverse Events v3.0. Results Median follow-up was 19.9 months (range, 4–138.9 months). Median LFFS was 20.8 months (95% CI, 15.1–24.3 months), and median OS was 20.9 months (95% CI, 17.1–27.2 months). Analysis of competing risks showed that, at 5 years, the probability of local failure as the first detected event was 17.1% (95% CI, 10.3–25.3%), the probability of distal failure as the first detected event was 17.9% (95% CI, 11–26.1%), and the probability of death without recording a failure was 50.7% (95% CI, 40.1–60.3%). Patients aged ≥ 70 years had higher probability of death without recording a failure within 5 years (55.5%; 95% CI, 41.7% −67.3%) than patients aged Conclusion Conformal radiation therapy is an effective and safe alternative to surgery for patients with stage I NSCLC. The results are limited because this is a single-institution observational study. Nevertheless, a large majority of patients remained free of local recurrence and without significant clinical toxicity.