Early Stage Pts Research Articles

A retrospective, descriptive analysis of a single-center experience with patients with metaplastic breast cancer.

e13173 Background: Metaplastic breast cancer (MpBC) is a subtype of breast cancer (BC) that is commonly triple negative (TN) and chemotherapy (CT) resistant. Despite CT resistance, standard of care therapy is administered according to receptor status, regardless of metaplastic histology. Randomized data regarding optimal treatment of this rare histologic subtype are lacking. Methods: This was a retrospective analysis of MpBC patients (pts) treated at Miami Cancer Institute (MCI) between 2017 and 2021. Pts were identified using COTA real-world Analytics, an analytics platform enabling rapid investigation of longitudinal real-world data. Pt demographics, pathologic characteristics, staging, treatment, and survival outcomes were collected and analyzed. Results: 46 pts with MpBC were identified. Median age at diagnosis was 60.5 years, and 83.7% were post-menopausal. 84.8% were White, 10.9% were Black, and 54.3% were Hispanic (all White). 80.4% (n=37) had TNBC (ER < 1% and HER2 negative by ASCO CAP guidelines). 13.0% of pts (n=6) had spindle cell morphology and these pts had numerically lower 5-yr overall survival (OS) compared to those with non-spindle cell morphology; (83.3%; 95% CI 53.5%-100% vs 91.7%; 95% CI 82.7%-100%). Of 21 pts (45.7%) who underwent germline genetic testing, a pathogenic variant was detected in 2 (9.5%); (BRCA2, PALBB2). At diagnosis, 93.5% (n=43) had early-stage disease, 2.2% (n=1) had de novo metastatic breast cancer (MBC), 2 were unknown. Of the early-stage pts, 12 (27.9%) had stage 1, 19 (44.1%) stage 2, and 6 (14.0%) stage 3 BC. Of 43 pts with treatment records available, 34 (79.1%) received CT and 20 (46.5%) received anthracycline-based CT in the early-stage setting. Of those who received CT, 50% (n=17) were treated in the neoadjuvant setting, and 2 pts (11.8%) achieved pathologic complete response (pCR). Median distant recurrence free survival (DRFS) and OS were not reached in the overall cohort. Two-year OS was 90.5% (95% CI 81.6%-99.4%) and two-year DRFS was 78.9% (95% CI 66.6%-91.2%). Of the 43 pts with early-stage MpBC at diagnosis, 11 (25.6%) developed MBC with lung as the most common site of metastasis. Only 15.2% (n=7) of tumors underwent somatic mutation analysis and 71.4% (5/7) were PIK3CA-mutated. Of the pts who developed MBC, median progression-free survival with cytotoxic chemotherapy was 9.0 months (95% CI 5.0-13.0). Conclusions: MpBC represents an aggressive phenotype of BC and optimal guidance regarding treatment is lacking. Consistent with published literature, a high proportion of MpBC in this study were TN (>80%). Our dataset represents the largest cohort to date of Hispanic pts with MpBC. Despite the poor outcomes historically reported with MpBC, median DRFS and OS were not reached in this cohort, with short follow up. Further prospective randomized data on the optimal treatment of MpBC is warranted and a larger multi-institution review is planned.

Clinical validation of a prognostic 7-marker IHC assay (7-IHC) in 382 patients (pts) with stage IB/IIA cutaneous melanoma (CM; MELARISK-001).

9572 Background: For pts with resected CM, adjuvant anti-PD1 is approved for AJCC v8 stage IIB–IV and BRAFi/MEKi for stage III–IV disease. Nonetheless, there remains a subgroup of pts within stages IB/IIA, which is at high risk of relapse and death, accounts for a considerable share of overall CM mortality (1), and lacks access to such therapy. Thus, adjuvant clinical trials in this “early”-stage setting may be merited. However, this subgroup is not detectable via AJCC staging alone. A prognostic 7-IHC assay that was developed (2), and analytically (3) and prospectively clinically validated (4) to identify stage IB/IIA pts at high risk of relapse and death, may be useful in selecting pts for adjuvant trials (2,4). Here, we report results from the MELARISK-001 multicenter study of that assay. Methods: MELARISK-001 enrolled consecutive pts diagnosed with stage IB/IIA CM from 2000–16, with available formalin-fixed paraffin-embedded primary melanoma. Specimens were analyzed by 7-IHC and pts classified as either high-risk or low-risk. Kaplan-Meier survival analysis and multivariate Cox regression analysis (MVA) against Breslow thickness, ulceration, age, and sex were performed. Results: MELARISK-001 included 382 pts; 247 (65%) pts in stage IB, 135 (35%) in stage IIA, all sentinel node-negative. Median Breslow thickness was 1.6 mm, median age 60Y. 212 pts (55%) were classified as 7-IHC high-risk, 170 pts (45%) as 7-IHC low-risk. Median follow-up for recurrence-free survival (RFS) was 90 mos, for melanoma-specific survival (MSS) 98 mos. Comparing 7-IHC high-risk vs low-risk pts, 5Y RFS was 79% vs 100%, 10Y RFS, 72% vs 99%, 5Y MSS, 92% vs 100%, and 10Y MSS 82% vs 100% (Table; logrank p<0.001). In MVA, 7-IHC risk class (hazard ratio [HR] 22.3; 95% CI 7.0–70.9), Breslow thickness (1.5; 1.1–1.9), and age (1.61; 1.0–2.5) were significant independent prognosticators of RFS. 7-IHC HR for MSS could not be computed due to 7-IHC's 100% sensitivity. Prognostic performance variables (95% CI) for RFS were: sensitivity 98% (91%–100%), positive predictive value 26% (20%–33%), negative predictive value 99% (97%–100%). Conclusions: 7-IHC risk categorization identified 98% of relapses and 100% of CM-related deaths in stage IB/IIA pts. In MVA, 7-IHC risk class was the strongest independent prognosticator of survival endpoints. 7-IHC high-risk pts have a relapse rate comparable to that of later-stage pts for whom adjuvant therapy is approved. MELARISK-001 provides further evidence supporting use of 7-IHC to identify early-stage pts at high risk of relapse and death who may benefit from adjuvant therapy trials. 1. Whiteman et al J Invest Dermatol 2015. 2. Meyer et al PloS ONE 2012. 3. Ziemer et al Diagnostics2023. 4. Meyer et al Eur J Cancer 2023. [Table: see text]

Neighborhood deprivation and race as predictors for cancer-related mortality in breast cancer.

174 Background: The socioeconomic status (SES) of individuals and their neighborhoods significantly influences their access to healthcare and the outcomes of their diseases. Inequities in opportunities, education, income, and infrastructure can contribute to worse prognosis and unfavorable outcomes for certain malignancies in areas with lower SES. We studied the association of race and neighborhood deprivation with the survival of breast cancer (BC). Methods: We linked the neighborhood deprivation index (NDI) data by the National Cancer Institute with the BC cohort from the SEER database diagnosed from 2010 to 2016. NDI score of each county is created using factor analysis to identify variables from 13 SES attributes from the dimensions: wealth/income, education, occupation, housing conditions. NDI was classified into quintiles (qn) (Q1 least deprived - Q5 most deprived). The association between NDI, race/ethnicity and overall survival (OS) and disease-specific survival (DSS) was analyzed using Cox multivariate regression model. All analyses were adjusted for age, race, grade, insurance, and treatments. Results: Most of the early-stage (ES) (N= 88, 572) and advanced-stage (AS) BC (N= 12336) patients (pts) were in the Q5 (ES- 27.4%, AS- 30%) qn. There was a predominance of racial minorities in Q5 and Q4 compared to Q1 qn (ES: Blacks 13-15% vs 8%, Hispanics 15% vs 6%; AS: Blacks 19-23% vs 14%, Hispanics 13% vs 7%, p<0.001). Uninsured ES pts were higher in Q5 compared to Q1 qn (2.2% vs 1.7%, p<0.001). Triple-negative BC was higher in Q5 compared to Q1 qn (ES: 14.5% vs 11.7%; AS: 16.5% vs 8.8%, p<0.001). ES BC pts who live in Q5 qn have inferior OS and DSS compared to Q1 qn, but this trend was not observed in advanced BC (Table 1). The interaction between NDI and race for OS and DSS was significant (p-interaction 0.0030 and 0.02 respectively) for the ES BC, but not for advanced BC (p-interaction 0.67 and 0.72 respectively). In ES BC, Whites residing in Q5 qn had higher overall (OS: HR= 1.23, 95% CI= 1.14-1.34, p<0.001) and disease-specific mortality (DSS: HR= 1.29, 95% CI= 1.15-1.45, p<0.001) compared to Whites from Q1 qn. However, this disparity in mortality based on the NDI was not observed for Blacks with ES BC (OS: HR= 1.2, 95% CI= 0.98-1.46, p=0.06, DSS: HR= 1.2, 95% CI= 0.94-1.55, p= 0.12). Conclusions: Early-stage BC pts from areas with worse NDI have poor OS and DSS. Blacks with early-stage BC have poor OS and DSS regardless of the SES of the neighborhoods. To improve their clinical outcomes, personalized treatments are needed for Black pts with early-stage BC. Targeted investments and policies should prioritize improving the SES of neighborhoods with high deprivation to reduce healthcare disparities and enhance BC outcomes.[Table: see text]

The value of preoperative imaging for identification of metastatic disease and prediction of adjuvant therapy in clinical early-stage cervical cancer

Objectives: The FIGO 2018 revised cervical cancer (CC) staging includes radiologic and pathologic data. No standardized guidelines exist guiding routine pre-operative (pre-op) imaging in patients (pts) with presumed early stage disease. Adjuvant chemoradiotherapy (CRT) increases morbidity, highlighting the need for improved preoperative risk assessment. Identification of clinical scenarios where imaging may be omitted is needed for cost effectiveness. We evaluated the ability of preoperative imaging to predict need for adjuvant CRT in pts with clinical early stage CC. Methods: This is a retrospective review of all pts with CC treated surgically at our institution between 2010 and 2017. All pts had negative imaging or no imaging prior to surgery. Demographic and clinicopathologic data was abstracted from medical records. Included pts had squamous, adenocarcinoma, or adenosquamous histology treated primarily with abdominal, laparoscopic, or robotic radical hysterectomy. Pts were assigned to 1 of 2 groups: those who had pre-op imaging with MRI, CT, or PET, and those who had none. Primary outcome was percentage of pts requiring adjuvant CRT by imaging group. Secondary outcomes were rate of nodal metastasis, five year overall survival (OS), progression free survival (PFS), and treatment-related morbidity. Statistical tests used were chi-square, logistic regression, and Kaplan-Meier survival analysis. Results: Of 147 pts, 104 underwent pre-op imaging and 43 did not. Demographics did not differ between groups, except for more self-pay pts (16% vs 39.5%, p=0.002) and pts with adenocarcinoma (31% vs 44%, p=0.047) receiving no pre-op imaging. Pre-op imaging was not associated with a significant difference in the rate of adjuvant CRT (40% vs 35%, p=0.49), or nodal disease (15% vs 11.6%, p=0.55). There was no difference in rate of recurrence (15% vs 11.6%, p=0.55), OS (91% vs 89%, p=0.92), or PFS (76% vs 83%, p=0.66). Groups had similar rates of morbidity (35% vs 30%, p=0.53). Small tumor size predicted negative lymph nodes; 4.1% of pts with tumors less than 2cm had positive nodes on final pathology (OR=0.11, p=0.028). Many pts with clinical early stage disease do not receive surgery; thus, additional data was abstracted for pts with clinical stages I-IIA2 CC who underwent primary CRT. Of 92 pts who might otherwise have been eligible for surgical treatment, 39 (43%) had imaging findings suspicious for lymph node metastasis. Download : Download high-res image (144KB) Download : Download full-size image Conclusions: Though limited by sample size, approximately 20% of all clinically early stage pts in our cohort were upstaged by imaging and were not offered primary surgery. Imaging appears to be useful for initial disposition to primary surgery vs. CRT, but does not predict the need for adjuvant CRT, recurrence risk, or survival in pts treated with primary surgery. In pts with small tumors, (

Abstract 6478: Circulating stromal cells in resectable pancreatic cancer is associated with high pathological stage and poor clinical outcomes

Abstract Background: Pancreatic cancer (PC) is a difficult malignancy to diagnosis and properly stage, with extremely poor outcomes even for patients (pts) with resectable disease. Cancer Associated Macrophage-Like cells (CAMLs) are a recently described circulating stromal cell common to the blood of cancer pts, whose presence and size (≥50µm) is a prognostic indicator of poor progression free survival (PFS) and overall survival (OS). However, the clinical value of CAMLs in PC has not been evaluated. We recruited treatment naïve PC pts referred for surgical resection before therapy induction to study CAML association to PFS/OS. We investigate whether a simple blood test can better identify pts with metastatic disease and act as a predictor for PFS and OS. Methods: 68 whole peripheral blood samples were drawn from untreated newly diagnosed PC pts referred for surgical resection prior to induction of therapy. Pts were recruited for a 2 year single blind prospective pilot study testing >5 CAMLs, CAML size (≥50µm), CA19-9, & CEA to clinical stage (CS), pathological stage (PS), and resectability. Pts were referred based on CS with resectable (R) (n=23), borderline resectable (BR) (n=27), or locally advanced (LA)/metastatic (M) (n=16); Stage I (n=50), Stage II (n=12), Stage III (n=0), Stage IV (n=3), and unknown stage (n=2). Blood samples (7.5mL) were taken prior to any neoadjuvant therapy. Blood was filtered by CellSieveTM filtration and CAMLs quantified. Analysis of CAMLs, protein markers, and resectability were used to evaluate PFS and OS significance by log-rank testing. Results: CAMLs were found in 92% (n=61/66) of pt samples and averaged 7.5 CAMLs/7.5 mL blood. In pts that received surgery, 22 pts were upstaged with PS; stage I (n=20), Stage II (n=16), Stage III (n=5), Stage IV (n=18), & 7 pts withdrawing prior to surgery. All early stage pts with ≥12 CAMLs (n=6) were upstaged to metastatic disease after surgical resection. CEA and CA19-9 serum were not significant for PFS (CEA HR 0.9, p=0.639 and CA19-9 HR 0.6, p=0.288). Further, R vs BR was not significant for PFS (HR 0.4 p=0.10) nor was BR vs LA (PFS HR=0.2 p=0.075), though in both comparisons resectability did trend non-significantly toward better PFS. >5 CAMLs was not a significant indicator of PFS or OS (PFS HR 0.6 p=0.175 and OS HR 1.0 p=0.902). However, ≥50µm CAMLs was a highly significant indicator of both PFS and OS (PFS HR 4.0 CI95% 2.0-7.8, p>0.001 and OS HR 2.3 CI95% 1.1-4.7, p=0.035). Conclusions: In PC, accurate staging/resectability is difficult, with 60% of pts being upstaged post-surgery and ~80% recurring in 2 years. There is a need for a better method to stratify surgical candidates that won't benefit from surgical resection. These data suggest that high CAML numbers at diagnosis identifies pts with metastatic disease which are less likely to benefit from surgical resection, and enlarged CAML size correlates to poorer PFS and OS. Citation Format: Kirby P. Gardner, Daniel L. Adams, Mohammed Aldakkak, Cha-Mei Tang, Susan Tsai. Circulating stromal cells in resectable pancreatic cancer is associated with high pathological stage and poor clinical outcomes [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6478.

Brief Pembrolizumab(PEM) Monotherapy Results in Complete and Near Complete Responses in the Majority of Untreated Patients with Classical Hodgkin Lymphoma (cHL): A Multicenter Phase 2 PET-Adapted Study of Sequential PEM and AVD

Background: Pembrolizumab (PEM), an anti-PD-1 antibody is approved for the treatment of relapsed/refractory cHL based on results of a pivotal trial demonstrating an overall response rate (ORR) of 69% and 22% complete response (CR) rate. To investigate the impact of PEM monotherapy in previously untreated patients (pts), we conducted a phase 2 clinical trial of sequential PEM followed by AVD chemotherapy for untreated cHL. Herein, we report the primary analysis of response (PET2) to single agent PEM in the frontline setting and the results of interim FDG18-FDG PET-CT (PET3) following AVD x 2. Methods: Pts > 18 years of age with newly diagnosed cHL stages I-IV, including early stage pts with at least 1 risk factor according to NCCN criteria, were eligible. Pts were treated sequentially with 3 cycles of PEM at 200 mg every 3 weeks followed by an interim PET-CT (PET2) for primary analysis. Subsequently, pts received 4-6 cycles of AVD chemotherapy based on initial stage, with PET-CT's repeated after 2 cycles of AVD and at the end of therapy. There was no consolidative radiotherapy. Response to single agent PEM was assessed by Lugano criteria and decline in total metabolic tumor volume (TMV). We hypothesized that PEM monotherapy prior to AVD would result in a CR rate of 50% and would reduce the frequency of PET-positivity on the interim PET (PET3) following 2 cycles of AVD compared to prior reports. Results: Thirty pts enrolled from September 2017- March 2019 with a data cutoff of June 14, 2019. The median age of participants was 30 (range, 21-77), and 4 pts were ages >60. Eleven were male, and the majority (83%) were Caucasian. Twelve pts had early unfavorable disease (6 each with elevated ESR, B-symptoms, and bulky mediastinal masses). Eighteen pts had advanced stage disease (5 Stage 3, 13 Stage 4; IPS score 3 -4 in 7, 1-2 in 10). Adverse risk factors for all 30 pts included bulky disease or large mediastinal mass (n=12), B-symptoms (n=14) and extranodal disease (n=16). Overall therapy was well tolerated and most adverse events were grade 1-2. Related events of any grade included hypertension (n=10), rash (n=9), anemia (n=8), infusion reactions (n=5), elevated liver function tests (n=5), arthralgias (n=4) and hypo- or hyperthyroidism (n=3). Grade 3 and 4 events (whether or not judged to be related) were as follows Gr3: lymphopenia (n=4), febrile neutropenia (n=3), neutropenia (n=3), hypertension (n=2), diarrhea (n=1), hyperglycemia (n=1), and hyponatremia (n=1); Gr 4: neutropenia (n=10); transaminitis (n=1); sepsis (n=1). The only Grade 3 or 4 immune-related adverse event was the one grade 4 transaminitis that resolved with steroid administration and a delay in therapy. Eleven of 30 pts achieved CR (37%; Deauville 1-3) following initial PEM monotherapy including three with large mediastinal masses. Four additional patients with bulky disease and four others had major reductions in disease following PEM monotherapy, but did not achieve CR by Lugano criteria (Fig 1A). To better characterize the depth of response, we quantitated the decline in total metabolic tumor volume (TMV) (Fig 1B). Eight of the 17 pts with < CR to PEM monotherapy for whom TMV could be analyzed had > 90% reduction in TMV. A single pt had an atypical response with clearance of original disease sites, but development of new sites that resolved after two cycles of AVD. Across all pts, the CR (Deauville 1-3) rate following 2 cycles of AVD, was 100%. Median follow up is 6.5 (range: 1 - 12) months among those who have completed therapy. No pt has required a change in treatment, relapsed, or progressed, with progression-free and overall survival rates both at 100%. Conclusion: Although the study did not meet its primary endpoint, our data demonstrate that 3 cycles of PEM is highly active in pts with newly diagnosed early unfavorable or advanced stage cHL. PEM monotherapy resulted in > 90% reduction in TMV in the majority of pts, and when followed by AVD x's 2, CR in 100%. No pts have experienced progression or change of therapy to date and treatment was overall very well tolerated. This radiotherapy- and bleomycin-free approach warrants further investigation in larger trials to confirm response rates and assess efficacy compared with other novel combinations in the frontline setting. Disclosures Pro: Takeda: Consultancy, Honoraria, Other: Travel Expenses; Celgene: Consultancy, Honoraria; Kyowa Hakka Kirin: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria, Other: Travel Expenses, Research Funding. Karmali:Gilead/Kite; Juno/Celgene: Consultancy, Speakers Bureau; Takeda, BMS: Other: Research Funding to Institution; Astrazeneca: Speakers Bureau. Gordon:Gilead: Other: Advisory Board; Juno/Celgene: Other: Advisory Board, Research Funding; Zylem LLC: Other: co-founder; research in nanoparticles in cancer; Bayer: Other: Advisory Board. Winter:Merck: Consultancy, Research Funding.

Abstract 3407: Development of a chemokine signature identifying dMMR patients (pts) with poor prognosis

Abstract BACKGROUND: Colorectal cancer (CRC) pts with deficient mismatch repair (dMMR) tumors have greater immune cell infiltration than pts with proficient (pMMR) tumors and is thought to be partly responsible for the robust response of dMMR pts to PD-1/PD-L1 blockade. Not all pts with dMMR tumors respond to PD-1 blockade. Therefore, we conducted exploratory analyses to understand the complexity of the tumor immune microenvironment. We developed a gene expression signature predictive for MMR status and tested it for associations with prognosis. METHODS: The one-sided Wilcoxon signed-rank test was used to select genes that were significantly over expressed in dMMR v pMMR tumors in a randomly selected discovery cohort in NSABP C-08 (N=500). Six chemokine genes were selected to build a predictive signature for MMR status because they were among the top 20 most significantly differentially expressed genes and were part of an immune chemo-attractant pathway. Addition of other genes did not improve prediction accuracy. The signature was tested in an independent cohort of C-08 (N=454) and in three other data sets: NSABP C-07 (N=1603), TCGA colon (N=619), and the NSABP Molecular Profiling Registry (MPR-1) (N=263). MPR-1 tissue samples were from primary tumors in pts who had recurred. Survival analyses were performed in C-07, C-08, and TCGA data, using Cox proportional hazards models. MMR status was determined for MPR-1 tissues with IHC using antibodies for MSH2, MLH1, PMS2, and MSH6. RESULTS: Higher chemokine scores (CKS) were significantly associated with dMMR tumors in all three independent data sets composed of treatment-naïve early-stage CRC pts (P<0.01). However, in primary tumors of dMMR pts who developed recurrent disease, CKS was non-significantly lower than pts with pMMR tumors (P=0.072). As expected, high CKS was associated with a better prognosis in all three data sets, composed of early-stage CRC and was a more robust prognostic indicator in all 3 datasets than was MMR status. When CKS was combined with MMR status in early-stage pts, pts with high CKS had a better prognosis than pts with low CKS regardless of MMR status. CONCLUSIONS: The CKS score has a strong association with MMR status in three different data sets and is associated with a better prognosis in early-stage CRC. In these pts, dMMR tumors had significantly higher CKS than pMMR tumors, but primary tumors from MPR1 pts who had all become metastatic showed a trend for association in the reverse direction. We conclude that dMMR primary tumors, which are likely to become metastatic, have a very different immune microenvironment than those that do not because the CKS is composed of chemo attractants for a variety of immune cells including NK, T, and B cells. Such a signature could be useful for prospectively identifying dMMR pts with early-stage CRC who are at high risk for relapse. SUPPORT: Genentech; sanofi; U10CA180868, UG1CA189867, U24CA196067, PA DoH; NSABP Foundation. Citation Format: Ying Wang, Rim S. Kim, Corey Lipchik, Ashok Srinivasan, Huichen Feng, Nan Song, Carmen J. Allegra, Patrick G. Gavin, Samuel A. Jacobs, Norman Wolmark, Peter C. Lucas, Katherine L. Pogue-Geile. Development of a chemokine signature identifying dMMR patients (pts) with poor prognosis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3407.

Multi-gene hereditary cancer testing, family history and prognosis in men with prostate cancer.

5073 Background: Multi-gene hereditary cancer testing is common in prostate cancer (PC). We assessed the frequency of pathogenic mutations (mt) and examined associations with family history (FH), cancer recurrence, and overall survival (OS). Methods: Men with clinically localized or metastatic PC consented to germline DNA testing using a validated panel of 30 genes associated with elevated risk for common cancers (Color Genomics). Chi-square test and Fisher’s exact test were used to compare pathogenic mt and clinical characteristics. The Kaplan-Meier method was used to evaluate the associations of mt status with PSA recurrence and OS. Results: 315 men (median age 69, range 38-89) were included; 140 (44.4%) with localized and 175 (55.6%) with metastatic PC. 215 (68.3%) reported a FH of any cancer in one and 101 (32.1%) in ≥ 2 first-degree relatives. Genomic testing detected 12.1% of pts possessed a pathogenic mt (39 mt in 38 pts): BRCA2 (n = 12, 3.8%), CHEK2 (n = 10, 3.2%), APC (n = 7, 2.2%), MUTYH (n = 3, 0.9%), ATM (n = 2, 0.6%), and 1 each (0.3%) with NBN, PALB2, RAD51D, MSH2, and PMS2. Variants of uncertain significance were detected in 52 (16.5%). Men with FH of prostate, breast, ovarian, pancreatic cancer or/and melanoma (182/315, 58%), were more likely to have BRCA2 pathogenic mt (P = 0.049). Median age of diagnosis was 55 with BRCA2 versus 62 in non-carriers (P = 0.01). No association was found between FH of Lynch-associated cancers (colon, endometrial, gastric) and mismatch repair gene variants. The prevalence of pathogenic mt was significantly higher with visceral metastases (P = 0.001) but did not differ between localized and advanced PC pts (P = 0.602). In men with clinically localized PC at diagnosis (n = 226, 71.7%), pathogenic mt were associated with a shorter time to recurrence (30 vs 53 months, P = 0.004). Men with advanced PC (n = 175) and germline APC or MUTYH mt had worse median OS (44 vs 318 months, P = 0.004). Conclusions: Our findings confirm the incidence of men with non-PC FH associated with DNA repair gene mt and support testing in early stage pts. Further, our data support the prognostic significance of germline genetic alterations which deserves study in a larger cohort of pts.

No Utility of Routine Laboratory Testing during Surveillance in Detecting Relapse in Patients with Classic Hodgkin Lymphoma in First Remission

Background: Classic Hodgkin lymphoma (CHL) is currently highly curable. Surveillance guidelines have been extrapolated largely from an era where sophisticated staging techniques were lacking, radiotherapy alone was the mainstay of therapy, and cure rates were lower. While the lack of a role for surveillance imaging for patients (pts) in complete remission (CR) at end of therapy (EOT) is well established, there is paucity of data regarding role of laboratory testing. Current NCCN guidelines recommend complete blood counts (CBC), chemistry panels, and erythrocyte sedimentation rate (ESR) “as clinically indicated,” while ESMO recommends testing be performed at each clinic visit. To assess the utility of surveillance laboratory testing in detecting relapse, we conducted a retrospective analysis.Methods: Newly diagnosed CHL pts, uniformly treated with the Stanford V regimen from 1998-2014 and in CR for at least 3 months, were identified from our institutional lymphoma database. Post EOT, follow up included; a clinic visit and laboratory tests (CBC, metabolic panel, ESR) every 2-3 months for years 1-2 and at 6-12 month intervals thereafter for years 3-5. Data at surveillance visits were abstracted from electronic medical records for up to five years after EOT. Laboratory tests categorized by CTCAE v4.03 as Grade 2 or higher were considered abnormal. Values outside the normal range were considered abnormal for non-categorized tests. Primary analysis included sensitivity, specificity, positive (PPV) and negative predictive values (NPV) of surveillance laboratory tests for predicting relapse in first three years after EOT. Secondary objectives included assessment of subsets of higher risk pts defined as early stage (ES) unfavorable [GHSG or EORTC criteria], early stage bulky disease, ESR > 30 mm/hr and advanced stage disease with an international prognostic score (IPS) 3-7.Results: 235 pts were identified. Characteristics included median age 32 (range 18-82), 92 (39.1%) ES favorable, 66 (28.0%) ES unfavorable, and 61 (25.9%) advanced stage. 16 (6.8%) of ES pts could not be classified as favorable or unfavorable due to missing data. 24 (10.2%) pts relapsed at a median time from EOT of 8 months (range 3.4-80.5). In the first three years after EOT among 234 pts (one did not undergo laboratory testing), the mean number of surveillance blood draws per patient was 7.1, range (1-13). These 1661 surveillance blood draws included 4684 individual laboratory tests, comprising 1609 CBCs, 1578 metabolic panels, and 1497 ESRs. 180 (77%) of pts had at least one abnormal test result over five years. None of the biopsies confirming relapses were prompted by any abnormal laboratory finding. Relapses were detected on basis of surveillance imaging (n=15, 65.2%), patient symptoms and/or abnormal physical exam (n= 8, 34.8%). The sensitivity of any surveillance laboratory test for detecting relapse within 3 years of EOT was 72.7% (95% CI: 49.8-89.3%), specificity 22.6% (17.2%-28.9%), yielding a PPV of 8.9% (95% CI: 7.0%-11.3%) and NPV of 88.9% (79%-94%). Similar results were seen in higher risk subsets (Table 1).Conclusion: Our study found no utility of routine surveillance laboratory testing in detecting relapse in pts with CR at EOT including higher risk subsets. Our results support modifications of current practice guidelines. DisclosuresLynch:Johnson Graffe Keay Moniz & Wick LLP: Consultancy; Rhizen Pharmaceuticals S.A.: Research Funding; Takeda Pharmaceuticals: Research Funding; T.G. Therapeutics: Research Funding; Incyte Corporation: Research Funding. Advani:Agensys: Other: Institutional Research Support; Celgene: Other: Institutional Research Support; Regeneron Pharmaceuticals, Inc.: Other: Institutional Research Support; Bayer Healthcare Pharmaceuticals: Other: Consultancy/Advisory Role; Bristol Myers Squibb: Other: Consultancy/Advisory role and Institutional Research Support; Millenium: Other: Institutional Research Support; Kura: Other: Institutional Research Support; Infinity: Other: Institutional Research Support; Kyowa: Other: Consulting/Advisory Role; Merck: Other: Institutional Research Support; Gilead/Kite: Other: Consultancy/Advisory Role; Forty Seven, Inc: Other: Institutional Research Support; Janssen Pharmaceutical: Other: Institutional Research Support; AstraZeneca: Other: Consultancy/Advisory Role; Roche/Genentech: Other: Consultancy/Advisory Role, Institutional Research Support; Takeda: Other: Consultancy/Advisory Role; Cell Medica: Other: Consultancy/Advisory Role; Autolus: Other: Consultancy/Advisory Role; Seattle Genetics: Other: Consultancy/Advisory role, Institutional Research Support; Pharmacyclics: Other: Institutional Research Support.

Abstract OT3-03-03: A prospective study of glycoprotein 88 (GP-88) blood test in healthy women undergoing screening for breast cancer (BC) with mammography (MM)

Abstract Background: Population based BC screening with XRAY mammography (MM) has been widely accepted as standard of care for women aged 40+ with average risk of developing BC. Sensitivity and specificity of MM is dependent on breast tissue density and up to ∼20% of BC are undetected by MM. The development of a dependable, low cost blood-based BC screening test to increase the sensitivity and specificity of currently existing BC screening methods is needed. Rationale: GP88 is expressed & secreted by BC cells & is not expressed by normal mammary epithelial cells, 2 retrospective randomized multi-site trials (a training study & a validation study of 300 cases each) demonstrated that elevated GP88 expression in estrogen positive (ER+) invasive BC was statistically correlated with a 4-fold increase in the risk of 5-yr BC recurrence. GP88 was an independent predictor of BC recurrence in multivariate analysis of other factors such as PR expression, tumor size, grade, lymph node status & stage. The quantitative GP88 EIA was developed to determine the amount of GP88 in biological fluids. The blood based EIA assay is highly specific for GP88 & both sensitive & linear over a wide dynamic range, i.e. detection of GP88 concentrations from 0.1 to 20ng/ml. A baseline GP-88 level of28.4 ± 5 ng/ml was established by us for healthy volunteers (HV). In BC pts a statistically significant increase of serum GP88 was observed in early stage pts (40.7 ± 16 ng/ml; p=0.007). Stratification of BC pts according to their clinical outcomes shows that pts having no evidence of disease (NED) have serum GP88 levels within the range of HV. These data suggest that pts with breast tumors express & secrete high levels of GP88. Objectives: 1. To determine prospectively GP-88 blood levels in HV at average risk of developing BC screened by MM & in women with recently biopsy-confirmed BC. 2. To establish the statistical distribution of GP88 serum levels in subjects by baseline BIRAD classification (1-6). 3. To determine if the initial GP88 level is predictive of change in BIRADS classification from baseline to 12-mos follow-up. 4. To determine if baseline GP88 level is predictive of the appearance of BC at 12 mos follow-up in HV who were cancer-free at study entry. Inclusion Criteria: Female, aged >=40 yrs old, presenting for screening or diagnostic MM or diagnostic workup and/or biopsy due to abnormal MM <= to 12 wks before study entry. Study procedures: Serum levels of GP88 in subjects with average BC risk factors will be measured prospectively at baseline; 3-6 mos & 6-12 mos & correlated with BIRADS reading of the screening MM, BIRADS 1-6; GP88 serum level will be correlated with pathologic results of breast biopsies performed on subjects with suspicious BIRADS (4 & 5) MM & final pathologically confirmed diagnosis of breast cancer as BIRADS 6. Study Progress: The study is ongoing; currently we have 308 subjects enrolled, the total number of subjects will be up to 725 & screened up to 1400. Study is UM IRB approved & is conducted at the University od Maryland Medical Center (UMMC) and UM Baltimore Washington Medical Center (BWMC). Funding is provided by Maryland Industry Partnership Grant (MIPS)& Avon Grant No. 02-2013-018. Citation Format: Tkaczuk KHR, Campassi C, Kesmodel S, Bellavance E, Rosenblatt P, Nichols E, Feigenberg SJ, Coughlin P, Drogula C, Urban B, Galandak J, Dromi S, Kuo L, Yue B, Hicks D, Serrero G. A prospective study of glycoprotein 88 (GP-88) blood test in healthy women undergoing screening for breast cancer (BC) with mammography (MM). [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr OT3-03-03.

Analysis of Early Mortality of Chronic Lymphocytic Leukemia (CLL) Patients Treated in US Practices in the Connect CLL® Registry

Analysis of Early Mortality of Chronic Lymphocytic Leukemia (CLL) Patients Treated in US Practices in the Connect CLL® Registry

Abstract OT2-4-01: A prospective study of glycoprotein 88 GP-88 blood test in healthy women with Gail model risk <=1.66 undergoing screening for breast cancer (BC) with mammography (MM)

Abstract Background: In the US, the majority of BC is diagnosed by screening XRAY mammography (MM) but up to ∼20% of BC are undetected by MM. The development of reliable, blood-based BC screening test to increase the sensitivity and specificity of currently existing BC screening methods such as MM. Rationale: GP88 is expressed & secreted by BC cells & is not expressed by normal mammary epithelial cells. Two retrospective randomized multi-site trials (one training study & one validating study with about 300 cases each) have demonstrated that elevated GP88 expression in ER+ IDC was statistically correlated with a 4-fold increase in the risk of 5-yr disease recurrence. Multivariate analysis showed that GP88 as a risk predictor was independent from PR expression, tumor size, grade, lymph node status & disease stage. The quantitative GP88 EIA developed to determine the amount of GP88 in biological fluids was developed at A&G. The assay is highly specific for GP88 & both sensitive & linear over a wide dynamic range, i.e. detection of GP88 concentrations from 0.1 to 20ng/ml. A baseline (28.4 ± 5 ng/ml) was established for healthy volunteers (HV). In BC pts a statistically significant increase of serum GP88 was seen for early stage pts (40.7 ± 16 ng/ml; p=0.007). Additionally, a stratification of pts according to their clinical outcomes shows that pts having no evidence of disease (NED) have serum GP88 levels within the range of HV. These data suggest that pts with breast tumors express & secrete high levels of GP88. Objectives: 1. To determine prospectively GP-88 blood levels in HV at average risk of developing BC who are undergoing screening MM & in women with biopsy-confirmed BC. 2. To establish the statistical distribution of GP88 serum levels in these women by baseline BIRAD classification (1-6). 3. To determine if baseline GP88 level is predictive of change in BIRADS classification from baseline to 12-mos follow-up. 4. To determine if baseline GP88 level is predictive of the appearance of BC at 12 mos follow-up in HV who were cancer-free at baseline. Inclusion Criteria: Female, aged >=40 yrs old, presenting to UM Breast Center for screening MM or for diagnostic MM or diagnostic workup and/or biopsy due to BIRADS 0 MM at an outside facility <= to 12 wks; Gail model assessment of risk of developing BC<=1.66. Study procedures: Serum levels of GP88 in subjects with average BC risk factors will be measured prospectively at baseline; 3-6 mos & 6-12 mos. & correlated with BIRADS reading of the screening MM, BIRADS 1 or 2 (benign), BIRADS 3 (short term MM follow-up) & BIRADS 4 or 5 (suspicious and tissue diagnosis needed immediately). GP88 blood level will be correlated with pathologic results of breast biopsies performed on subjects with suspicious BIRADS (4 & 5) MM & final pathologically confirmed diagnosis of BC as BIRADS 6. Statistical Considerations The total number of pts will be 725 & screened up to 1400 subjects. Study is UM IRB approved & just started accruing. Funding is provided by Maryland Industry Partnership Grant (MIPS)& Avon Grant No. 02-2013-018. Contact: Ntait@umm.edu. Citation Format: Katherine HR Tkaczuk, Cristina Campassi, Susan Kesmodel, Emily Bellavance, John Olson, Elizabeth Nichols, Steven J Feigenberg, Binbin Yue, David Hicks, Ginette Serrero. A prospective study of glycoprotein 88 GP-88 blood test in healthy women with Gail model risk <=1.66 undergoing screening for breast cancer (BC) with mammography (MM) [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr OT2-4-01.

The natural history of veterans with colorectal cancer (CRC) over a period of 10 years.

e14626 Background: We studied demographics, comorbidity, treatment, and other predictors of survival for patients (pts) with CRC to define the natural history Methods: In an IRB-approved protocol, we reviewed the records of pts diagnosed with CRC at a VA medical center from 01/01/2003 to 12/31/2012. Demographics, grade, stage, ECOG performance status (PS), CEA, Hemoglobin (HGB), and Albumin (ALB) at diagnosis, diagnosis by screening, cause of death, recurrence, survival, and treatment were studied. Comorbidity was assessed with the Charlson Comorbidity Index (CCI). Statistical analyses were performed with SAS 9.2 and Stata v11.0. Results: There were 296 pts with 294 men and 2 women. The median (M) age was 69 years (44-90). There were 4 (1.4%) pts with stage 0, 95 (33%) with stage I, 72(25%) stage II, 56(20%) with stage III, 57(20%) with stage IV CRC. The percentage of pts diagnosed by screening varied inversely with stage, from 43% (41 pts) for stage I, 22% (16 pts) for stage II, 20% (11 pts) for stage III, to 2% (1 pt) for stage IV (p<.001). M survival (MS) was 1,126 days (4-4004) with 143 pts (48%) deceased. Survival varied inversely with stage, with median survivals of 1,548 days for stage I, 1,557 days for stage II, 1011 days for stage III, and 413 days stage IV CRC pts (p<.001). Surgery was done in 232 pts (78%), chemotherapy in 102 pts (33%). Median ECOG PS was 1(0-4) and Charlson 4.1 (0.8-13.9). Death from CRC occurred in 5 (5%) pts with stage I, 17 (24%) pts with stage II, 15 (27%) pts with stage III, and 37(65%) pts with stage IV CRC (p<.001). Recurrence was experienced by 22 early stage pts (9%). For all stages, M HGB was 12.4 (6.5-16.7) gm/dl, CEA 2.5 (0.2-2071) ng/ml, and albumin 4.0 (2.1-4.8) gm/dl at diagnosis. By univariate analysis, stage (p<.001), HGB (p<.001), grade (p<.001), ECOG PS (p<.001), CCI (p<.001), CEA (p<.001), surgery (p<.001), chemotherapy (.049), and diagnosis by screening (p<.001) were predictive. In multivariate model, stage, Charlson, chemotherapy, ECOG PS, and albumin levels were predictive. Conclusions: In this study population, screening colonoscopy and chemotherapy are showing an effect on survival. Confirmation in large patient samples is planned. Supported by NJCCR

Fifteen-year experience of all patients (pts) with small bowel adenocarcinoma (SBA), treated in a specialized gastrointestinal (GI) oncology unit: Royal Marsden (RM) experience.

316 Background: Small bowel adenocarcinoma (SBA) is a rare tumour with poor prognosis. There is paucity of published literature due to rarity of disease; we conducted this retrospective study to determine the clinical course and outcome along with prognostic factors in both early and later stage SBA. Methods: Clinical characteristics and outcomes of all pts treated consecutively in the GI Unit RM, 1996-2011 were recorded. The study endpoints were relapse free survival (RFS), progression free survival (PFS), and overall survival (OS), in early stage pts (G1) and in pts with advanced disease (presentation or relapse with un-resectable disease=G2). In G2 response rate (RR) to chemotherapy was determined. In both groups association to baseline prognostic factors were sought by performing Cox regression univariate analysis (UVA). Results: Eighty four pts with SBA were treated 1996-2011. A total of 48 presented with early stage disease (G1). In G1 (58.3% males; mean age, 57 years), 44/48 pts underwent R0 resection; 21 received adjuvant chemotherapy. RFS, PFS and OS in this group were 29.6 [95% confidence interval (CI) 3.3-55.9], 31.1 (CI=8.0-54.3) and 42.9 (CI=0-94.9) months (m), with median follow up of 76.4 m. Poor histological differentiation (p=0.025), abnormal CEA at presentation (P=0.082), and lymphovascular invasion (p=0.003) were prognostic of OS. G2 comprised of 36 pts with un-resectable disease along with 23 from G1 who subsequently relapsed [G2 (n=59); 52.5% males; mean age, 59 years]; 54 pts with metastatic and 5 with locally advanced disease; 78% received first-line chemotherapy. Overall RR of pts who received chemotherapy was 50%. OS and PFS were 12.8 (CI =8.4-17.2) and 8.8 (CI=5.5-12.3) m respectively; 1-year survival was 60.9% vs. 27.3% (no chemotherapy) (p=0.042). Abnormal albumin (0.041), platelet count (p=0.007) and CEA (p=0.025) were prognostic of OS in the chemotherapy group; doublet (18/41) versus triplet (23/41) chemotherapy were not prognostic (p=0.185). Conclusions: Pts with SBA and metastatic disease may derive benefit from systemic chemotherapy; prospective clinical trials are required to evaluate this further.

Aspirin use and survival outcomes in patients (pts) with PIK3CA mutant colorectal cancer (CRC).

3598 Background: Aspirin has an established role in preventing CRC. Recent data suggests aspirin use may also benefit a subset of pts diagnosed with CRC. In one series, the authors identified PIK3CA mutations as a potential predictive biomarker for aspirin use, reporting that PIK3CA mutant CRC pts receiving aspirin had superior cancer specific survival (CSS) compared to those not receiving aspirin (HR 0.18, p&lt;0.001), whereas in pts with wildtype PIK3CA, aspirin had no survival impact. Our study aims to confirm the survival benefit associated with aspirin use in pts with PIK3CA mutant CRC. Methods: A cohort of CRC pts with PIK3CA mutations (Sanger sequencing) was identified. Prospectively collected clinicopathological, treatment and outcome data was available. Aspirin use was confirmed by chart review. CSS and overall survival (OS) analyses were conducted using Cox proportional hazards in univariate and multivariate settings. Recurrence free survival (RFS) analyses were limited to early stage CRC pts (Stage A-C). Statistical differences in 5-year CSS (5YS) rates were calculated using Fisher’s exact test. Results: From a cohort of 1,019 CRC pts with known PIK3CA mutation status, 121 (12%) harbored PIK3CA mutations. Of these, 112 (92%) had aspirin usage data available: 27 (24%) pts used aspirin, 85 (76%) did not. In the aspirin group, there were 22 (81%) early stage and 5 (19%) metastatic CRC pts; in the no-aspirin group, there were 59 (69%) early stage and 26 (31%) metastatic CRC pts. In univariate analyses, aspirin use was not associated with superior CSS (HR 0.57, p=0.21), OS (HR 0.83, p=0.57), or RFS (HR 0.72, p=0.57). In multivariate analyses, aspirin use was not associated with improved OS (HR 1.07, p=0.86), CSS (HR 1.04, p=0.94) or RFS (HR 0.54, p=0.34). In 69 (62%) pts with mature follow-up, there was a trend towards superior 5YS for aspirin users (69% v 42%, p=0.09), but this may reflect imbalances in stage at diagnosis. Conclusions: Our study was unable to confirm the recently reported survival benefit associated with aspirin use in pts with PIK3CA mutant CRC. Given the small numbers of pts, a modest survival benefit associated with aspirin use cannot be excluded. Analyses in an expanded cohort of early stage pts are underway.

6501 ORAL Observations of Hepatocellular Carcinoma (HCC) Management Patterns From the Multinational HCC BRIDGE Study – First Overall Analysis of the European Cohort

Results: 1571 pts comprise the safety population; 54% were BCLC stage C. Sor treatment duration was longer in the early stage (A/B) pts (Table). BCLC stage did not appear to determine Sor dosing. Early stage pts received more prior locoregional treatment (LRT), and more prior and concomitant transarterial chemoembolization (TACE). Early and advanced staged pts had comparable percentages of drug-related AEs (DRAEs) and drug-related serious AEs (DRSAEs). In the ITT population (N = 1612), the preliminary median overall survival (OS) was 9.1 months (mos). OS according to BCLC stage at study entry was 13.6mos in stage A pts (n = 117), 12.6mos in B (n = 311), 7.9 mos in C (n = 877) and 3.4mos in D (n = 93). Conclusions: Preliminary 2 IA results indicate that Sor is used across all BCLC stages. The Sor safety profile was generally similar across BCLC stages. Treatment duration and preliminary OS data of Sor in the BCLC subset supports the robust nature of the GIDEON study data.

Chronic lymphocytic leukemia (CLL) patients at a VA medical center: Comorbidity and survival.

6602 Background: To determine whether co morbidity indices predict survival in chronic lymphocytic leukemia (CLL) patients (pts). Methods: In an IRB-approved protocol, we reviewed the records of 107 pts diagnosed with CLL at a VA Medical Center from January 2001 to May 2010. Records were reviewed for demographic, clinical, pathological data. ECOG PS, Veteran status (V-Vietnam) treatment status and survival were tabulated. Comorbidity was assessed using Charlson Comorbidity Index (CCI), The Kaplan-Feinstein Index (KFI), the Cumulative Illness Rating Scale (CIRS) and the VA comorbidity Index. We preformed Cox regression analysis to determine predictors of survival. Results: There were 107 pts with a median (M) age of 72 yo (48-94). The M hemoglobin(Hgb) was 13.6 g/dl (7.3-16.7), M WBC 15.4 K/cmm (2.8-289), M platelets 201 K/cmm (50-1077), M blood urea nitrogen (BUN) 19 mg/dL (1-88), M creatinine (Cr) 1.1 mg/dl (0.4-3.6), M lactate dehydrogenase (LDH) 177 IU/L (105-1635), M beta 2-microglobulin 2.3 mg/dl (0.9-8.9) and M BMI 28.7 kg/m2 (15.06-43.1). Early stage pts numbered 94 (88%). The M CCI was 4 (0.8-8.7), M KFI 1 (0-3), M CIRS 15 3 (1-6), M CIRS 16 7 (1-14), M CIRS 17 2 (1-4.5), M VA comorbidity 3 (0-8).106 pts had a PS 0-2. By immunophenotyping 42 ptswere CD38/ ZAP-70 positive. 65 pts (75 %) non-V, 30 (28%) V. The M survival was 1290 days (4-6715). In univariate analysis age (P<0.01), stage (p< 0.03), BMI (P 0.05) beta 2-microglobulin (p< 0.0001), Hgb (p<0.005), WBC (p<0.0004), BUN (p<0.0006), Cr (p<0.0002), LDH (p<0.0033,), second malignancy (P<0.0436) and V (p<0.04) were significant predictors of survival. In the multivariate analysis, B2-Microglobulin was the only significant independent predictor of survival (p < 0.0172). Comorbidity indices were not predictors of survival. Conclusions: Co morbidity indices were not able to able predict survival in this group as many patients were early stage pts with good performance status.

Racial/ethnic disparities in lung cancer stage of diagnosis and survival

8080 Background: There are differences in the treatment outcome of non-small cell lung cancer (NSCLC) patients (pts) between non-Hispanic whites (NHW) and African Americans (AA). Little is known regarding the outcomes of Hispanics (H). Methods: Registry data on 2,696 pts with NSCLC treated during 1999–2006 was obtained. The objective of the study was to evaluate differences in NSCLC survival according to different ethnicities. Chi-square was used to compare distribution of tumor stage. Survival curves were compared using log-rank test for each of the tumor stages. Adjusted hazard ratios (AHR) and 95% confidence intervals (95% CI) were reported based on the results of a multivariate Cox regression model for overall survival (OS) with adjustment for gender, age at diagnosis, and race. Results: Most pts had stage III/IV at diagnosis; majority of the AA or HW presented in advanced stage compared with NHW. Significantly higher proportions of AA and H were diagnosed with stage IV compared to NHW ( Table ). Mean age at diagnosis was 62 yrs (AA 58, H 60, and NHW 66yrs) and it was significantly different among the 3 groups (one-way ANOVA, p&lt;0.0001). AA and H have significantly shorter stage-specific median survival for early stage compared to that in NHW ( Table ). In pts with advanced stages the pattern was similar: AA and H have a significantly shorter median survival than that in NHW. In early-stage pts, significant predictors for OS from multivariate Cox regression model were female gender (AHR=0.65; p&lt;0.001), AA (NHW as the referent group; AHR=2.67; p&lt;0.0001), and H (NHW as the referent group; AHR=2.01; p&lt;0.0001). In late-stage pts, significant predictors for OS were female gender (AHR=0.79; p=0.0002), AA (NHW as the referent group; AHR=1.53; p&lt;0.0001, and H (NHW as the referent group; AHR=1.28; p=0.0006). Conclusions: NHW pts had better OS than H and AA; we will evaluate whether gene expression profiles or presence of EGFR overexpression have an impact on racial/ethnic disparities in the outcome of NSCLC. [Table: see text] No significant financial relationships to disclose.

The clinical significance of chromosome 17p deletion in chronic lymphocytic leukemia: A study of 180 consecutive patients

7035 Background: Preliminary data indicate that the presence of chromosome 17p deletion (17p-) by fluorescent in-situ hybridization (FISH) in chronic lymphocytic leukemia (CLL) may be uniquely associated with resistance to standard therapy and dismal survival. Methods: In order to determine the significance of 17p- in CLL in a large cohort, we analyzed the clinical course and treatment outcome of 180 consecutive 17p- CLL pts managed at our center. Results: *PT CHARACTERISTICS. 56 (31%) pts were untreated (UNT), and 124 (69%) had relapsed / refractory (RR) disease. RR pts were more likely to have impaired performance status (PS) (p=0.01), Rai stage 3/4 disease (p<0.01), and elevated β2m (p=0.07). *PROGRESSION TO FIRST THERAPY. For UNT patients with asymptomatic CLL, the 18 month progression to therapy risk was 46%, with 18 pts remaining treatment-free at 7 - 47 months. The 17p- clone was lost spontaneously in 2 pts on follow-up. *RESPONSE TO INITIAL THERAPY. Six UNT pts received rituximab, with 4 (67%) objective responses (OR) and 2 (33%) complete clinical responses (CCR). Twenty-nine UNT pts received fludarabine & rituximab therapy, with 22 (76%) OR and 14 (48%) CCR. All CCR pts tested by FISH showed elimination of their 17p- clone. *RESPONSE TO SALVAGE THERAPY. Results by treatment categories were: rituximab, OR 27% & CCR 0%; alemtuzumab containing regimens, OR 50% & CCR 19%; fludarabine & rituximab therapy, OR 44% & CCR 0%; intensive chemotherapy, OR 20% & CCR 4%; investigational agents, OR 10% & CCR 0%. SURVIVAL. UNT pts had a favorable survival substantially better than that reported in previous studies (median not reached, 2 yr survival 81%), whereas RR pts had a median survival of only 14 months (p<0.0001). Among RR pts, those with a normal performance status (n=35) had a favorable 2 year survival of 78%, with no deaths between 12 and 46 months. Conclusions: The outcome of 17p- CLL is not necessarily dismal. Half of early stage pts will not require therapy at a follow-up of up to 4 years. The 17p- clone may be lost spontaneously or be eliminated by effective therapy. Rituximab containing regimens are highly effective particularly in the frontline setting. No significant financial relationships to disclose.

Co-Expression of CD38 with Zap-70 Is Predictive of Treatment Intervention in Patients (pts) with Early Stage Chronic Lymphocytic Leukemia (CLL).

Introduction: CLL is a heterogeneous disease where advanced stage pts have compromised survival despite treatment, while early stage pts may not require any intervention. Based on current NCI-WG guidelines, most early stage (Rai stage 0 and 1) CLL pts do not require treatment until the development of progressive or symptomatic disease, though eventually over 70% of CLL pts will receive therapy. Currently, there are no markers to predict this subset of pts. Several markers of adverse prognosis, including ZAP-70+, CD38+, un-mutated IgVH gene and cytogenetic aberrations (17p-, 11q-, +12) have been identified. These markers are primarily studied in context of aggressive clinical course and survival outcome. The value of these markers to predict the necessity of treatment intervention in early stage CLL pts has not yet been completely evaluated. We investigated the expression pattern of Zap-70 and CD38 to examine their predictive value in identifying early stage CLL pts who will require therapeutic intervention.Methods & Results: 93 CLL pts were evaluated since 2002 at our institution. Flow cytometry was used to determine Zap-70 and CD38 expression on CD19+ CLL cells obtained from peripheral blood. Pts were considered to be positive for Zap-70 and CD38 expression if ≥ 20% and ≥ 30% of the cell stained for these proteins, respectively. For the Zap-70 analysis we used similar methodology as described by Crespo et al.1 Thirty-six (19M, 17F) pts had limited stage CLL, based on Rai staging criterion. Median age was 65 years (range 43–86) with stage 0 or 1 observed in 14 and 22 pts, respectively. Median time from diagnosis is 2 years (range <1–17). Increased expression of either Zap-70 or CD38 was seen in 22 and 7 pts, respectively. Five (14%) pts were positive for both. All (100%) pts with concurrent increased expression of Zap-70 and CD38 required treatment (p=0.003), as compared to 30% of patients requiring treatment that were negative for both proteins or positive for only protein. Among the pts that required treatment 4 had 13q- and one had trisomy 12 on cytogenetic analysis.Conclusion: Our study, demonstrates for the first time, the clinical utility of CD38 and Zap-70 co-expression in determining the probability of treatment intervention in early stage CLL pts. Although the number of pts studied is small, our findings highlight a potentially important use of these markers in the management of early-stage CLL pts. These observations warrant validation in a larger cohort of pts early stage CLL pts as well as correlation with other prognostic markers such as cytogenetic and IgVH gene mutational status.