Results of a pilot study to increase adherence to ASCO G-CSF recommendations at community clinics.

19 Background: An ASCO 2012 Choosing Wisely recommendation cautions against the use of primary prophylactic colony stimulating factors (PP-CSF) for chemotherapy regimens with <20% risk of febrile neutropenia (FN). Our pilot aimed to test the feasibility and impact of using academic detailing and an automated CSF ordering system on CSF prescribing at 6 regional community oncology clinics in Washington State. Methods: The intervention was 1) academic detailing for oncologists during a regular staff meeting and 2) reconfiguring ordering systems to show FN risk and PP-CSF recommendation in a passive alert and include or exclude PP-CSF standing orders with >20% or ≤20% FN risk regimens, respectively. Clinic tumor registries were queried for patients with stage II-IV breast, non-small cell lung, or colorectal cancer starting first-line chemotherapy pre- or post-intervention. PP-CSF use, FN risk factors, and chemotherapy regimens matching the study protocol were manually abstracted. A regimen order was coded as adherent if it is low (<10%) FN risk without PP-CSF use; intermediate (10-20%) FN risk without PP-CSF use or FN risk factors are present; or high (>20%) FN risk with PP-CSF use. Results: The intervention was successfully implemented at 4 out of 6 participating clinics. The remaining 2 clinics were transitioning from paper to electronic orders in 1-2 years or using a system managed by a non-participating hospital. Adherence across the four implementing clinics increased after the intervention. *=p<0.05 Implementation of the CSF order-entry intervention was successful across a variety of ordering systems, including paper-based systems. Overall, while adherence prior to the intervention was high for these clinics, the order entry systems significantly improved adherence. Population characteristics and data availability may account for variation in adherence. Conclusions: An intervention with both academic detailing and ordering system presets may help increase adherence to Choosing Wisely recommendations. [Table: see text]

How to Manage the Obese Patient With Cancer

Purpose Obesity (body mass index [BMI] ≥ 30 kg/m2) is common among patients with cancer. We reviewed management issues in the obese patient with cancer, focusing on how obesity influences treatment selection (including chemotherapy dosing), affects chemotherapy toxicity and surgical complications, and might be a treatment effect modifier. Methods The majority of evidence is drawn from observational studies and secondary analyses of trial data, typically analyzed in N × 3 BMI categories (normal weight, overweight, and obese) matrix structures. We propose a methodological framework for interpretation focusing on sample size and composition, nonlinearity, and unmeasured confounding. Results There is a common perception that obesity is associated with increased treatment-related toxicity. Accordingly, cytotoxic chemotherapy dose reduction is common in patients with elevated BMI. Contrary to this, there is some evidence that full dosing in obese patients does not result in increased toxicity. However, these data are from a limited number of regimens, and fail to fully capture cytotoxic drug pharmacodynamics and pharmacokinetic variability in obese patients. Among patients undergoing surgery, there is evidence that elevated BMI is associated with increased perioperative mortality and increased rates of infectious complications. A novel finding is that these relationships hold after surgery for malignancy, but not for benign indications. There are biologic plausibilities that obesity might be an effect modifier of treatment, but supporting evidence from clinical studies is inconsistent. Conclusion In line with the ASCO 2012 guidelines, chemotherapy dosing is probably best performed using actual body weight in obese patients. However, specific regimens known to be associated with increased toxicity in this group should be used with caution. There is no guidance on dose for obese patients treated with biologic agents. Currently, there are no specific recommendations for the surgical management of the obese patient with cancer.

Treatment Patterns and Outcomes Among Patients Who Initiate Therapy for Relapsed/Refractory Multiple Myeloma (RRMM) in Routine Care

Background Multiple myeloma (MM) is the second most common hematologic cancer in adults. Novel agents, including thalidomide, bortezomib, and lenalidomide, have led to improved overall survival (OS) in MM, but the disease remains generally incurable with the majority of patients inevitably relapsing after front-line therapy (FLT). In a multi-center observational analysis of 383 MM patients (treated between 2007 and 2010), median progression-free survival (PFS) and OS from treatment after first relapse were 13 and 35 mos (Durie ASCO 2012; abs 8095). According to NCCN guidelines, retreatment with primary therapy may be considered if relapse occurs more than 6 mos after discontinuation of primary therapy, otherwise a switch in regimen is recommended (NCCN MM Guidelines v.4 2015). Treatment patterns and outcomes in RRMM remain to be elucidated. We examined treatment use, retreatment and therapy switch patterns, and outcomes among patients with RRMM initiating second-line therapy (SLT). Methods We conducted a retrospective cohort study using a large, national, US claims database. Adult patients with MM who initiated cancer-specific systemic therapy between Jan 2008 and Feb 2014, without evidence of transplantation, were identified. Newly diagnosed MM patients were followed from the first claim for MM. Continuous enrollment in the health plan for 12 mos pre-diagnosis (with no claim for MM) through at least the start of SLT was required. FLT began with the first claim for MM-directed systemic cancer therapy. Unique agents administered within 90 days of FLT initiation constituted a regimen. Continuation of FLT regimen (or part thereof) or monotherapy within 3 mos of the end of the initial regimen was considered part of FLT. SLT after first relapse for RRMM was identified accordingly: 1) a treatment gap >6 mos between end of FLT and start of a second regimen (retreatment or switch), 2) start of a follow up regimen (retreatment) with a treatment gap of >3 and up to 6 mos after end of FLT, or 3) a switch to another drug combination after FLT regimen. Analyses were conducted from the first claim for SLT. SLT ended at the earliest of start of a new drug, death, or end of study period (Feb 2014). MM drug combinations were based on all unique MM systemic therapy agents received within start and end date for SLT. Vital status information was ascertained from the Social Security Death Index. Results We identified 249 patients with RRMM receiving SLT. Approximately half (49%) were male; 29% and 43% were aged 65-74 and 75 years or older, respectively. SLT regimens were: lenalidomide±dexamethasone (R±d, n=70 [28%]), bortezomib±d (V±d 61 [25%]), other (58 [23%], see Table 1), V+other (25 [10%]), R+other (14 [6%]), V-cyclophosphamide±d (VC±d, 13 [5%]), and VR±d (8 [3%]). The switch and retreatment patterns from FLT to SLT are depicted in Table 1. Among patients with treatment-free interval (TFI, time from end of FLT to start of SLT) >6 mos (n=64) vs ≤6 mos (n=185), 18 (28%) vs 17 (9%) were retreated with the same primary regimen in SLT, respectively. Median duration of SLT was 6.3 (95% CI: 5.6, 6.9) mos (Fig. 1) and of FLT was 6.8 (95% CI: 6.0, 7.8) mos. The 1-year OS probability from initiation of SLT for RRMM was 82% (95% CI: 76%, 86%). Conclusions Among patients with RRMM treated in the USA, V- and R-based regimens were the most common at first relapse. The vast majority of patients (91%) with TFI ≤6 mos switched therapy at time of relapse in concordance with NCCN guidelines. The relatively short duration of SLT (median 6.3 mos) in this study compared with PFS from treatment after first relapse (median 13 mos, Durie ASCO 2012; abs 8095) in a cohort of patients with RRMM, suggests that the majority of patients discontinue SLT prior to disease progression. These findings highlight the need for newer SLT regimens that are effective and more sustainable compared with current treatment choices. Table 1. Switch and Retreatment Patterns from FLT to SLT among Patients with RRMM. SLT V-based R-based VR-based Other FLT V-based 41 (41%) 49 (58%) 4 (50%) 15 (26%) R-based 33 (33%) 13 (15%) 2 (25%) 12 (21%) VR-based 9 (9%) 6 (7%) 2 (25%) 19 (33%) Other* 16 (16%) 16 (19%) 0 (0%) 12 (21%) Total 99 (100%) 84 (100%) 8 (100%) 58 (100%) *Other combinations consisted of a subset of the following agents: cyclophosphamide, melphalan, vincristine, doxorubicin, interferon-alpha, pomalidomide, thalidomide, carfilzomib, dexamethasone, prednisone Disclosures Romanus: Millennium Pharmaceuticals Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited: Employment. Jhaveri:Sanofi: Equity Ownership; Takeda Pharmaceutical Company Limited: Equity Ownership; Millennium Pharmaceuticals Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited: Employment. Labotka:Millennium Pharmaceuticals, Inc., Cambridge, MA, USA, a wholly owned subsidiary of Takeda Pharmaceutical Company Limited: Employment. Henk:Optum (a consulting firm retained by Takeda to conduct the reasearch pertaining to this abstract): Employment. Seal:Millennium Pharmaceuticals Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited: Employment.

Updated use of first-line chemotherapy for advanced pancreatic cancer: FOLFIRINOX versus gemcitabine.

6607 Background: Pancreatic cancer (PC) is the fourth leading cause of death in the United States. It is estimated that 45,220 patients will be diagnosed in 2013 and 38,460 will die (Siegel, CA Cancer J Clin 2013). Gemcitabine has long been the standard of care chemotherapy. Recent advances in treatment created a combination regimen (oxaliplatin, irinotecan, leucovorin, fluorouracil [FOLFIRINOX]) for patients with good Karnofsky performance status (PS) (Conroy, NEJM 2011). This retrospective analysis was conducted as an update to results reported at ASCO 2012 (Ginsburg Arlen, JCO 2012) to evaluate characteristics and overall survival (OS) of patients receiving FOLFIRINOX and gemcitabine-based treatments in a large outpatient community setting. This is the largest study describing FOLFIRINOX patients to date. Methods: Patients with advanced PC treated within The US Oncology Network entered into the iKnowMed (iKM) database between June 2010 and November 2012 were included. Patterns of treatment were characterized by the median age at diagnosis, sex, PS, and first-line metastatic chemotherapy prescribed. The primary endpoints of the analysis were OS and uptake of FOLFIRINOX within the network. Results: Compared to ASCO 2012 results, 1,000 additional patients were identified in iKM. Of the 1,714 total patients, 24% received FOLFIRINOX (up from 13% in 2012) and 76% gemcitabine-based therapy (87% in 2012). Increased utilization of FOLFIRINOX for patients with good PS began in June 2010. For all patients (55% male), the median age at diagnosis was 67 years and the majority (85%) had a PS of 70% or greater. The OS was significantly longer for FOLFIRINOX (9.6 mos) versus gemcitabine (6.3 mos) (p<0.0001). This held true for PS of 70% or greater patient given FOLFIRINOX (9.6 mos) versus gemcitabine (7 mos) (p<0.0001). Conclusions: Utilization of FOLFIRINOX has continued to expand after the publication of phase III trials. Our data in a community setting supports a survival advantage for FOLFIRINOX. Although the magnitude of benefit may be smaller in the community, we agree that FOLFIRINOX should become a standard of care for good PS patients.

Impact of the trifunctional antibody catumaxomab followed by systemic chemotherapy (CTX) on overall survival (OS): Results of a subgroup analysis in patients (pts) with relapsed ovarian cancer (OC) and malignant ascites (MA).

e16547 Background: MA can compromise QoL in pts with OC. The safety and efficacy of catumaxomab in pts with MA due to EpCAM+ carcinomas where standard therapy was not available, not effective or no longer feasible was confirmed in a phase III trial comparing a 3-h intraperitoneal (i.p.) infusion with and without prednisolone premedication (Sehouli, ASCO 2012). Pts with OC who received subsequent CTX after treatment with catumaxomab (post-catumaxomab CTX) were analysed separately. Methods: OC pts in both treatment arms were pooled. The efficacy parameters time to next puncture (TTPu), puncture-free survival (PuFS) and OS were evaluated in relation to post-catumaxomab CTX. Results: 42/109 (39%) pts with MA due to OC received post-catumaxomab CTX, which included a platinum-containing regimen in 19 pts (17%). 21/109 (19%) received >1 post-catumaxomab CTX regimen. Pts with any post-catumaxomab CTX compared with those without CTX had significantly prolonged OS (median 273 vs 81 d, HR 0.24, p<0.0001) and PuFS (median 138 vs 43 d, HR 0.46, p = 0.0002). The difference in TTPu was not significant (223 vs 110 d, HR 0.68, p = 0.1788). Pts with >1 post-catumaxomab CTX compared with 1 post-catumaxomab CTX had significantly prolonged OS (480 vs 167 d, HR 0.20, p < 0.0001). The difference between the groups with 1 and >1 post-catumaxomab CTX in PuFS (153 vs 123 d, HR 0.64, p = 0.1804) and TTPu (153 vs 169 d, HR 1.52, p = 0.3600) was not significant. Pts who received platinum-containing post-catumaxomab CTX had significantly prolonged OS compared with those who received post-catumaxomab CTX without platinum (462 vs 169 d, HR 0.32, p = 0.0026). The difference between the groups with and without platinum in PuFS (153 vs 123 d, HR 0.76, p = 0.4448) and TTPu (153 vs 229 d, HR 1.68, p = 0.2373) was not significant. Conclusions: The data indicate that pts with MA due to OC who are able to receive CTX after catumaxomab treatment have significantly prolonged survival compared with pts who could not receive CTX. Further trials have been initiated to identify the best patient population to receive i.p. catumaxomab followed by subsequent CTX.

The role of social exposure to smoking on smoking cessation in adult cancer survivors.

9536 Background: We previously described a strong inverse relationship between social smoking exposures (at home, spousal and with peers) and smoking cessation in lung cancer, with adjusted odds ratios (aOR) of 3-8 (Eng et al, ASCO 2012, abst 9032). In the current analysis, we evaluated whether these associations hold true in adult cancers in general, particularly cancers not traditionally known to have smoking as a risk factor. Methods: 616 cancer survivors across multiple cancer sites were surveyed on their smoking, alcohol, and physical activity habits before and at various times after cancer diagnosis. Social smoking exposures were documented. Multivariate logistic regression models evaluated the association of each variable with change in each habit after diagnosis adjusted for significant socio-demographic and clinico-pathological covariates. Results: Median follow-up after diagnosis was 26 months. 15% had breast cancers; 15% gastrointestinal; 20% genitourinary-gynecological; 24% haematological; 36% other. Among current smokers at diagnosis, 56% quit after diagnosis; no ex- or never-smoker restarted. Patients without secondary home smoking exposure were significantly more likely to quit smoking than those with home exposures (aOR=9.5, 95% CI [2.4-37.8]). Similar results were seen in patients with non-smoking spouses versus smoking spouses (aOR=3.7 [1.0-13.4]), and with lack of peer smoke exposure (aOR=3.7 [1.3-10.7]). 63% patients who quit did so in the 1 year period surrounding the diagnosis date (6-months pre or post diagnosis). In comparison, first and second-hand smoking exposures did not affect other modifiable behaviours such as alcohol or physical activity. Patient awareness of quality of life and survival benefits of smoking cessation and receiving smoking cessation counselling were not associated with improved smoking cessation. Conclusions: Secondary smoking exposures are associated with lack of smoking cessation in adult cancer survivors, even in cancers not traditionally linked to smoking. Being diagnosed with cancer may be an important `teachable moment` to help patients quit, but results are strongly influenced by the surrounding social exposure to smoking. PS and GL contributed equally.

Quantitative analysis of the impact of deepness of response on post-progression survival time following first-line treatment in patients with mCRC.

3630 Background: The extent of tumor shrinkage in patients (pts) receiving chemotherapy with or without monoclonal antibodies is prognostic for PFS and OS. The ‘Deepness of response (DpR)’ concept aims to relate tumor shrinkage to post-progression survival (PPS). If tumor shrinkage occurs, DpR is the percentage of shrinkage observed at the nadir compared with baseline. DpR is 0 for no change and negative if the tumor load increases. Longest diameter (LD) based on RECIST or a calculated tumor volume (ASCO GI 2012 #635) can be used to quantify the tumor load. A joint model was presented (ASCO GI 2012 #580, ASCO 2012 #3603) which allows the prediction of a pt PPS time based on DpR. Methods: Based on data from the randomized CRYSTAL and OPUS trials, 4 treatment regimens (FOLFIRI +/- cetuximab and FOLFOX4 +/- cetuximab) were studied. A joint model was used to quantify individual changes in tumor size over time and to relate these changes to PFS and OS. Relationships between baseline tumor load and DpR and PPS were studied. A Spearman correlation was used to study the relationship between DpR and PPS for KRAS wild-type (wt) pts with progressive disease. Results: Results are reported using LD-based measures for 841 pts with KRAS wt tumors and imaging data. The 348 pts treated with FOLFIRI alone had a median DpR of 33.3% (interquartile range [IR]: 8.0%, 58.0%) while the 315 pts treated with FOLFIRI + cetuximab had a significantly higher median DpR of 50.9% (IR: 18.4%, 78.6%), p<0.0001. The 96 pts treated with FOLFOX4 alone had a median DpR of 30.7% (IR: 4.0%, 55.9%) and the 82 pts treated with FOLFOX4 + cetuximab had a significantly higher median DpR of 57.9% (IR: 24.0%, 92.9%), p=0.0008. Correlation between DpR and PPS for pts with documented progression is statistically significant in each treatment group for both LD-based and volume-based measurements (p<0.0001 for the CRYSTAL study and p<0.005 for the OPUS study). Conclusions: Our results emphasize the value of DpR as a new efficacy outcome measure for clinical trials. The tumor-shrinking capacity of cetuximab was shown to be associated with its ability to prolong PPS.

SPIO-enhanced MR imaging at 3T for the detection of metastases in sentinel nodes in patients with breast cancer.

1107 Background: We previously demonstrated the usefulness of SPIO-enhanced MR imaging at 1.5T for the detection of metastases in sentinel nodes localized by computed tomography lymphography (CT-LG) in patients with breast cancer (Ann Surg Oncol 2011, ASCO 2012). The aim of this study was to evaluate the accuracy of MR imaging at 3T with SPIO enhancement for the detection of metastases in sentinel nodes. Methods: This study included 60 patients with breast cancer and clinically negative nodes. Sentinel nodes identified by CT-LG were evaluated prospectively using SPIO-enhanced MR imaging at 3T. A node was considered non-metastatic if it showed a homogenous low signal intensity and metastatic if the entire node or a focal area did not show a low signal intensity on MR imaging. Sentinel nodes located by CT-LG were removed, and imaging results and histopathological findings were compared. Results: The mean patient age was 54.2 years (range, 33-78). Sentinel nodes were identified by CT-LG successfully in 59 (98.3%) of 60 patients. The mean number of sentinel nodes identified by CT-LG was 1.43 (range, 1-3). All 16 patients with positive sentinel nodes definitively diagnosed by pathology demonstrated metastases on SPIO-enhanced MR imaging. Five (31.3%) of them had micrometastases. Forty-one of 43 patients with negative sentinel nodes definitively diagnosed by pathology were non-metastatic on imaging studies. The sensitivity, specificity and accuracy of MR imaging for the diagnosis of sentinel node metastases were 100%, 95.3%, and 96.6%, respectively. No adverse events were associated with either CT or MR imaging. Conclusions: SPIO-enhanced MR imaging at 3T is useful for accurate diagnosis of sentinel node metastases, and therefore sentinel node biopsy may be avoided for most patients with breast cancer.

ER as a predictor of early breast cancer (EBC) outcomes in patients.

590 Background: Some ER-negative (ER-) breast cancers express low levels of estrogen receptors and approximately 12% express androgen receptors (Traina, T, et al. ASCO, 2012). Whether young premenopausal women (age <40) with ER- breast cancer (BC) who are more likely to retain ovarian function after adjuvant chemotherapy have a worse outcome than older women with ER- disease has not been widely investigated. Methods: We analyzed 2 adjuvant US Oncology BC studies: 99-016, 1830 BC patients randomized to doxorubicin/cyclophosphamide (AC)→Paclitaxel (P) (AC/P) vs AP→weekly P (no cyclophosphamide [C]) (AP/wP); and 01-062, 2611 patients randomized to AC→docetaxel (T) vs AC→T plus capecitabine (XT). ER+ patients received standard endocrine therapy following chemotherapy. Five-year DFS results did not show significant differences between the treatment arms on either study. The outcomes were analyzed for 5-year DFS by age ≤40yrs and >40yrs and by ER status. Results: In the two studies combined, ER- patients ≤40 had a superior DFS (84%) than ER- patients >40 (80%), while ER+ patients ≤40 had a worse 5-yr DFS (83%) than ER+ patients >40 (89%), although these findings were of borderline significance (see Table below). In 99-016, omitting C did not adversely affect outcomes in either age group, regardless of ER status. Conclusions: We did not observe worse outcomes in ER- patients ≤40 years compared to those >40 years in 2 US Oncology adjuvant chemotherapy trials, suggesting no adverse impact of assumed greater ovarian function following adjuvant chemotherapy in patients ≤40yrs. ER+ patients ≤40 had a worse DFS than ER+ patients >40. Omitting C in ER- patients ≤40 or >40 did not adversely affect outcome. [Table: see text]

Hematologic safety of Ra-223 dichloride (Ra-223) in castration-resistant prostate cancer (CRPC) patients with bone metastases from the phase III ALSYMPCA trial.

5060 Background: In the updated analysis of the ALSYMPCA study, Ra-223 significantly improved median survival by 3.6 mo vs placebo (Pbo) in 921 CRPC patients (pts) with bone mets (HR = 0.695; 95% CI, 0.581-0.832; p = 0.00007) and had a highly favorable safety profile (Parker et al. ASCO 2012). The hematologic safety profile and results from a post hoc analysis assessing prognostic factors for changes in hematologic parameters in ALSYMPCA pts are presented. Methods: Pts were randomized 2:1 to 6 injections (inj) of Ra-223 (50 kBq/kg IV) q4wk or matching Pbo and stratified by prior docetaxel (D) use, total alkaline phosphatase (tALP), and current bisphosphonate use.Multivariate regression analysis was performed to explore the relationship of 6 baseline factors (Table) with maximum % change of hematologic parameters from baseline up to 24 wk on treatment (tx). Results: The updated analysis included 901 pts (safety population; Ra-223, n = 600; Pbo, n = 301). Overall grade 3/4 AEs were similar between Ra-223 and Pbo groups, with neutropenia in 2% and 1%, thrombocytopenia in 6% and 2%, and anemia in 13% and 13% of Ra-223 and Pbo groups, respectively. Tx with Ra-223, prior D use, extent of disease (EOD) > 6 bone mets, and tALP ≥ 220 U/L, but not current bisphosphonate use, were associated with decreases from baseline in hemoglobin (Hb), neutrophils, or platelets; prior EBRT to bone for pain was associated with an increase. The significance of the relationship between baseline factors and changes in hematologic parameters is summarized in the Table. Conclusions: Overall, there was a low risk for hematologic AEs with Ra-223 tx in CRPC pts with bone mets. The strongest prognostic factors for decreases in neutrophils and platelets were Ra-223 tx and prior D use. Baseline tALP ≥ 220 U/L was a strong predictor of decrease in Hb. Clinical trial information: NCT00699751. [Table: see text]