Virtual Library

Start Your Search

P. Bonomi



Author of

  • +

    O02 - NSCLC - Combined Modality Therapy I (ID 111)

    • Event: WCLC 2013
    • Type: Oral Abstract Session
    • Track: Combined Modality
    • Presentations: 1
    • +

      O02.06 - Predictors of trimodality therapy use and overall survival in patients with stage III non-small cell lung cancer (NSCLC) in the National Cancer Database (ID 1736)

      10:30 - 12:00  |  Author(s): P. Bonomi

      • Abstract
      • Presentation
      • Slides

      Background
      The optimal locoregional therapy for stage III non-small cell lung cancer is controversial, with definitive chemoradiotherapy (CRT) and trimodality therapy (chemoradiotherapy followed by surgery, TMT) serving as competing strategies. The implementation of TMT and resultant survival outcomes in routine United States clinical practice has not been closely examined. In this study, we used the National Cancer Database (NCDB) to determine predictors of TMT and compare overall survival (OS) among a large cohort of stage III NSCLC patients treated with CRT or TMT.

      Methods
      Patients included were stage III NSCLC patients who received concurrent CRT with or without subsequent surgical resection at Commission on Cancer-accredited programs between 1998 and 2010; survival data were available for patients treated through 2005. High-volume (HV) center was defined as the upper decile. Per NCDB coding, treatment centers were stratified into academic, comprehensive, and non-comprehensive community cancer center (CCC). Logistic regression was used for univariable analyses, and multivariable models were prepared using stepwise selection to determine demographic, clinical and non-clinical predictors of TMT use and overall survival. Propensity score matching was used to estimate treatment effect and to minimize the effect of confounding variables.

      Results
      The overall cohort consisted of 49,534 patients, 25,679 of whom also had available survival data. Trimodality therapy was delivered in 7.8% of patients. Multivariable clinical predictors of TMT included: white race (OR 1.36), younger age (lowest [LQ] vs highest [HQ] quartile, OR 4.31), high school education or higher (HQ vs LQ, OR 1.37), stage IIIA vs IIIB, (OR 3.02) and squamous histology (OR 1.26). Non-clinical variables associated with TMT included: early treatment era (first vs last 3 years, OR 1.30), private insurance (OR 1.53), increasing distance from the treatment center (HQ vs LQ, OR 1.81), geography (Northeast vs. other, OR 1.35), treatment at academic research programs (ARP) over comprehensive CCC or non-comprehensive CCC (OR 1.53 and 2.01, respectively), and HV institution (OR 1.26). The median, 3- and 5-year OS were 12.3 months, 17% and 9.6%, respectively. Univariable comparison between CRT and TMT showed OS benefit with TMT (median 26.1 vs 11.7 months, log rank p<0.001). On MVA, stage IIIA (hazard ratio, HR, 0.85), squamous histology (HR 0.97), young age (HR 0.74) , female gender (HR 0.86), non-white race (HR 0.90), higher income (HQ vs LQ HR 0.76), farther distance from treatment center (HR 0.93), more recent treatment era (HR 0.80 vs. first 2 time periods) and TMT (HR 0.49) were significantly associated with improved OS. ARP was significantly associated with superior OS (HR 0.92 vs. CCP) if TMT was not in the Cox model. After propensity matching, TMT was still associated with improved OS (HR 0.62, p<0.001).

      Conclusion
      The use of TMT was strongly associated with markers of higher socioeconomic status and treatment at high-volume and academic centers. These data further support a significant survival benefit in patients undergoing TMT, and treatment at academic centers may improve OS via increased use of TMT.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    P1.06 - Poster Session 1 - Prognostic and Predictive Biomarkers (ID 161)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Biology
    • Presentations: 2
    • +

      P1.06-051 - Development of a serum biomarker panel predicting clinical outcome of chemotherapy with pemetrexed in patients with NSCLC (ID 3355)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      Pemetrexed disodium is a novel folate antimetabolite approved for first-line treatment in combination with a platinum doublet, for second-line treatment as a single agent and, more recently, as maintenance treatment after first-line chemotherapy in patients with non-squamous non-small cell lung cancer (NSCLC). Circulating factors associated with folate metabolism and/or phenotypic plasticity (e.g. the epithelial-to-mesenchymal transition (EMT)) may have predictive value in selecting advanced NSCLC for first-line pemetrexed. The objective of this study was to identify serum biomarkers capable of predicting improved outcomes for pemetrexed added to first-line platinum based chemotherapy relative to standard platinum doublet.

      Methods
      Pretreatment serum from a total of 72 patients with non-squamous stage IV NSCLC was evaluated with 76 biomarkers using Luminex immunobead assays. Patients were treated either with platinum combined with pemetrexed (P: n= 26) or with other agents (O; n=51) at the discretion of the treating physician. Patients were evaluated for disease progression using RECIST criteria. Biomarker data was processed using Ingenuity Pathway Analysis (IPA) Suite to identify interactions with folate metabolism. Cox Proportional Hazard (PH) regression model was used to assess the association between H-scores and progression-free/overall survival (PFS/OS) distribution estimated by the Kaplan-Meier method. PH interaction model was used to capture the differential effects of the biomarkers on the O vs. P treatment groups.

      Results
      Univariate PH regression analysis identified 10 biomarkers that were negatively associated (p<0.05) with progression-free survival (PFS) in either the O (sTNF-RI, sTNF-RII, Tenascin C, sIL-2Rα, spg130, sIL-6R, CA-125, and CA 19-9) or the P subgroups (total PSA, amphiregulin). Four other biomarkers (MMP-1, MMP-2, sVEGFR2, and PDGF-B) were all significantly (p<0.05) positively associated with PFS in the P group. Similarly, seven biomarkers were strongly negatively associated (p<0.01) with overall survival (OS) in the O group, including osteopontin, sTNF-RI, sTNF-RII, CA 15-3, sIL-2Rα, CYFRA 21.1, and IL-6; whereas the P group possessed both negative (osteopontin and amphiregulin) and positive (sVEGFR2, MMP-1, MMP-2, and sRAGE) associations (P<0.05) with OS. In our assessment of differential association with PFS, we found two serum biomarkers (PSA (total) and amphiregulin) with significant positive interaction terms, thus indicating differentially increased hazard of progression in the P group with higher level of the biomarker. MMP-1, HGF, and Tenascin C, sVEGFR2 were similarly noted to have significant negative interaction terms for PFS. Evaluations of the differential associations with respect to OS, demonstrated five biomarkers with significant (MMP-1, MMP-2, sVEGFR2, sTNF-RI, and Tenascin C; p≤0.05) and three strongly associated (osteopontin, HGF, s-IL-6R; p≤0.01) negative interaction terms, demonstrating a decreased hazard of progression in the P group.

      Conclusion
      Serum biomarkers with potential predictive (PFS, OS) value for selecting patients most likely to benefit from pemetrexed have been identified. Pathway analysis demonstrates interactions of biomarker candidates identified with folate metabolism. This study is currently being expanded with additional front-line patients (P=90; n=56) from our institutional archives to further evaluate their potential predictive value.

    • +

      P1.06-052 - Biomarkers of phenotypic plasticity associated with clinical outcomes in patients with locally-advanced NSCLC treated with chemoradiation with and without surgery. (ID 3019)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      Thoracic chemoradiotherapy (CRT) with or without surgery (S) is the standard-of-care in management of stage III NSCLC. However, it appears that plateau has been reached. New treatment strategies are needed. The objective of this retrospective study was to evaluate the relationships between patient outcomes and expression of biomarkers associated with either the epithelial-to-mesenchymal transition, or EMT ( E-cadherin and vimentin), or a lung cancer “stem-cell” phenotype (CD133), DNA repair enzyme (ERCC1), and cell survival/apoptosis (BCL-2, surviving and PTEN) in attempt to identify new therapeutic strategies.

      Methods
      Stage III NSCLC pts who were treated with chest radiation (40-65Gy) and concurrently with platinum doublet and who had sufficient pretreatment tissue were included in this study. Surgical pts received 40-45 Gy of radiation preoperatively and non-surgical patients received 60-65 Gy. Immunohistochemistry was used to detect nuclear and cytoplasmic expression of ERCC1, bcl-2, survivin, PTEN, vimentin, E-cadherin, and CD133. Scores were calculated using the Allred scoring system. The log-rank tests used to evaluate progression free survival (PFS) and overall survival (OS) with Kaplan-Meier plots used to plot group characteristics.

      Results
      A total of 119 patients receiving chemoradiotherapy with adequate tumor specimens for analysis were enrolled in this study; 61 had definitive chemoradiation whereas 58 had pulmonary resection after chemoradiation. Patients (n=79) with low nuclear survivin immunostaining (score ≤6) had significantly improved PFS as compared to those patients (n=34) with higher expression levels (14.1 vs. 10.5 months, p=0.042). Patients (n=72) with a cytoplasmic vimentin score ≤ 5 had superior PFS than the with higher expression levels (n=33) (13.17 vs. 9.99 months, p=0.045). High nuclear ERCC1 values (n=72) were associated with a worse OS than those patients (n=44) with low immunostaining (22.7 vs. 59.1 months; p=0.023). Patients with low cytoplasmic E-cadherin (n=25) had a significantly better OS than those patients (n=85) with immunostaining scores (62.6 vs. 24.6 months, p=0.036). The cytoplasmic vimentin/ E-cadherin ratio provided the most impressive separation of cohort performance with high V/E ratios being associated with a poor PFS (12.6 vs. 3.1 months; score ratio 10 cutoff; p=0.00073). No significant associations with cytoplasmic CD133 were observed in this cohort for either PFS or OS.

      Conclusion
      The association of inferior overall survival in locally-advanced NSCLC patients whose tumors express high ERCC1, high cytoplasmic E-cadherin (which is associated with mesenchymal phenotype and lower adherence of cells which are able to metastasize easier), and lower progression free survival with high survivin and high vimentin/ E-cadherin ratio suggests that combining inhibitors of survivin, DNA repair, EMT pathways might improve outcomes in molecularly defined subset of stage III NSCLC patients.

  • +

    P1.10 - Poster Session 1 - Chemotherapy (ID 204)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Medical Oncology
    • Presentations: 3
    • +

      P1.10-004 - Exploratory Subset Analysis in African Americans from the PointBreak Study (Randomized Phase 3 Pemetrexed + Carboplatin + Bevacizumab Followed by Maintenance Pemetrexed + Bevacizumab Versus Paclitaxel + Carboplatin + Bevacizumab followed by Maintenance Bevacizumab in Patients with Stage IIIB/ IV   Nonsquamous Non-Small Cell Lung Cancer) (ID 249)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      African Americans (AA) have a higher rate of lung cancer than Caucasians per 100,000 population (74.7 versus 64.4) but are underrepresented in randomized clinical trials. In the PointBreak study, AAs were enrolled at the same rate as the US incidence of non-small cell lung cancer (NSCLC) for AAs in 2011 (13%; now 15% in 2013). Post-hoc analyses of the AA subgroup were conducted from this study to evaluate efficacy and safety, as well as these outcomes by treatment center.

      Methods
      All patients (N=939) were chemonaive with an ECOG performance status (PS) of 0/1. AAs were analyzed against Caucasians for efficacy/safety in the pemetrexed (Pem) arm only. Subgroup analyses of AAs alone were conducted for efficacy/safety (between arm comparisons) as well as academic versus community settings (pooled two treatment arms). Hazard ratios and p-values were derived from a multivariate Cox-PH model, adjusting for disease stage, gender, PS and measurable/nonmeasurable disease. Response rates and adverse events (AEs) were compared using the exact test.

      Results
      Patients had stage IIIb (with pleural effusion)/IV nonsquamous NSCLC, according to AJCC edition 6. There were 94 AAs (42 = Pem arm; 52 = Paclitaxel [Pac] arm) and 805 Caucasians in the treated population. Demographics were statistically comparable between AAs and Caucasians in the Pem Arm, respectively: 62%/53% male, 71%/53% ≤65 years, 98%/89% ever smokers, 81%/90% Stage IV disease, with ECOG PS of 0/1, 33%/67%, versus 44%/56%. Median OS was similar between AAs and Caucasians in the Pem arm at 12.4 versus 12.3 months. Median PFS for AAs and Caucasians was 4.6 months versus 6.0 months (HR 1.229 (0.864 – 1.749; p=0.251). Overall response rate (ORR) for AAs was higher, though not statistically, at 38.1% versus 33.3% (p=0.607). Efficacy among AAs was fairly similar with median OS for Pem arm at 12.4 versus 13.7 months for Pac arm (p=0.1208). Median PFS by arm was 4.6 versus 5.1 months (p=0.6699). ORR among AAs was 38% in both arms. AAs showed a heavy trend (80%) for enrollment in community centers (n=74) versus academic center (n=20). Efficacy among AAs by setting showed a higher median OS at academic sites at 16.5 months versus 11.4 months, p=0.1906 (HR=0.6605; 95% CI: 0.355 – 1.229). Median PFS was also higher at academic sites at 6.9 months versus 4.6 months, p=0.9149 (HR =0.9690; 95% CI: 0.544 – 1.726). Drug-related grade 3/4 AEs in the Pem arm showed higher percentages for Caucasians with the exception of neutropenia, as follows: anemia (7.3%/15.9%), thrombocytopenia (9.8%/25.5%), fatigue (4.9%/11.5%), neutropenia (31.7%/25.3%), febrile neutropenia (0%/1.6%). Among AAs in the Pem/Pac Arm respectively, drug-related grade 3/4 AEs were: anemia (7.3%/0%), thrombocytopenia (9.8%/4.0%), fatigue (4.9%/4.0%), neutropenia (31.7%/44.0%), and febrile neutropenia (0%/4.0%).

      Conclusion
      There were no significant differences between AAs and Caucasians for OS, PFS, and ORR. Among AAs, median OS was not superior for either arm. PFS and OS were similar for academic and community settings among AAs. Caucasians had a significantly higher incidence of Grade 3/4 thrombocytopenia (p=0.0217), but this should be interpreted with caution due to the sample size for the AAs.

    • +

      P1.10-034 - Weight Gain as a Prognostic Factor on Patient Outcomes<br /> In Advanced, Nonsquamous, Non-small Cell Lung Cancer (ID 1905)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      Patients with advanced/metastatic non-small cell lung cancer (NSCLC) have a poor prognosis and low survival rates. One of the first notable symptoms of advanced lung cancer is unexplained weight loss. We evaluated weight gain (> 5% post baseline), as an early prognostic factor for clinical outcome, in advanced nonsquamous, NSCLC patients.

      Methods
      This retrospective analysis reports on three randomized phase III studies with survival and response data from a total of 2301 advanced, nonsquamous NSCLC patients who received pemetrexed or other chemotherapy plus a platinum or targeted agent, as first-line therapy. Body weight was recorded before and after treatment by each study’s schedule. Baseline weight was defined as the last non-missing weight measure before first treatment. Post baseline weight was defined as the maximum weight measured after starting treatment. Patients were analyzed using log-rank test and adjusted Cox modeling to assess the relationship between weight gain and overall survival (OS) and progression-free survival (PFS). Logistic regression was used to assess the association between baseline covariates and post-baseline weight gain.

      Results
      Patients were a mean age of 61 years (range 26 – 86) and most were of Caucasian descent (77.0%). A majority of patients had adenocarcinoma (73.8%), were male (59.8%) with an ECOG performance status (PS) of 0/1/2 (38.5%/60.2%/1.4%). Many patients were smokers or former smokers (55.7%) with Stage IV disease (83.1 %), according to the American Joint Committee on Cancer, editions. 5/6 and had an average weight at baseline of 71.4 kg. A total of 421 (18.3%) patients had a >5% increase in weight (>5% subgroup) after baseline with a statistically significant increase in OS and PFS. Median OS was 16.7 months for patients in the >5% subgroup versus 10.7 months for patients who gained <5% weight (< 5% subgroup; [n=1880]; p<0.001). PFS was 6.9 months for the >5% subgroup versus 4.8 months for <5% subgroup; p<0.001). Differences in overall response rate (ORR = CR + PR) and disease control rate (DCR = CR + PR + SD) were also significant. ORR was 50.8% for >5% subgroup versus 25.4% for < 5% subgroup (p<0.001). DCR was 91.5% for >5% subgroup and 63.6% for <5% subgroup (p<0.001). Cox modeling revealed patients in the >5% subgroup had significantly longer survival (HR=0.56, [95% CI 0.49-0.64]; p<0.001) than patients with <5% subgroup, after adjusting for baseline age (<65 versus 65), sex, ECOG PS (0 versus 1/2), histology (adenocarcinoma versus others), and study. Similar significant results were also found for PFS. Logistic regression indicated a significant association between weight gain and age. More patients aged <65 had a >5% weight gain (p<0.001).

      Conclusion
      This exploratory analysis showed that substantial weight gain (>5%) occurred after initiation of platinum-based chemotherapy in approximately 20% of advanced/metastatic, nonsquamous NSCLC patients. There was a positive correlation between weight gain and improved, OS, PFS and response in patients treated in these phase III studies.

    • +

      P1.10-036 - Prolongation of overall survival in non-small cell lung cancer before and after Food and Drug Administration (FDA) approval of active second line agents: a meta-analysis (ID 2051)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      Overall survival (OS) may be confounded by subsequent treatments in advanced NSCLC. To explore the potential effect of newer second-line and subsequent therapies on the results of clinical trials, a systematic literature review was conducted to examine OS and progression free (PFS) data from first line treatment trials published prior to and after the emergence of targeted therapies and newer FDA-approved second-line agents.

      Methods
      The PubMed database was searched for phase III first line trials in NSCLC over two 2 time frames: 1998 to 2005 and 2006 to 2011. Trials were included if platinum-based doublets were either the investigational or control regimen in previously chemotherapy-naive patients and if published in the English language. Bayesian statistical methods were used in random effects meta-analysis and meta-regression.

      Results
      13 clinical trials were included from the 1998-2005 and 17 trials from 2006-2011. The median percentage of women and adenocarcinoma included in the two time frames were 28 and 33, and 44 and 52 percentages, respectively. The difference in median survivals for the early and recent time frames (9.35 and 11.03 mo) in the treatment group was found to be significantly different (95% interval being shifted away from 0, Fig 1). However, the difference in the control group (9.74 and 10.59 mo) was not significant. The temporal trend in median survival over the time scale was further examined using meta regression with time as a covariate. The time trends or regression slopes for both the treatment (0.37, 95% interval 0.14 to 0.61) and the control (0.25, 0.06 to 0.43) arms were found to be significantly positive suggesting increasing trends in median survival over time. Only 18 of the 30 trials had PFS data with only 5 from the 1998-2005 timeframe. No significant difference in PFS was found over the two time periods.Figure 1

      Conclusion
      These analyses suggest longer survival in more recently treated NSCLC patients without a significant difference in PFS. Although this observation may have been influenced by the inclusion of more women and adenocarcinoma patients and improvements in supportive care, in the later time period more effective second and third line therapy may have contributed to longer OS.

  • +

    P1.18 - Poster Session 1 - Pathology (ID 175)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Pathology
    • Presentations: 1
    • +

      P1.18-012 - Thyroid transcription factor (TTF-1) negative lung adenocarcinomas will be wild type for epidermal growth factor receptor (EGFR) mutations. (ID 2036)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      TTF-1 is expressed in approximately 70% of adenocarcinomas (ACs) of the lung. EGFR mutations are present in 13-15% of unselected patients with AC in the United States and national guidelines suggest initiating first line EGFR tyrosine kinase inhibitors in this population. Both high TTF1 expression and EGFR mutations are associated with terminal respiratory unit (TRU) type ACs, female sex, never-smoking status and longer survival. We hypothesized that TTF-1 negative AC would have a high probability of being negative for EGFR mutations.

      Methods
      Microdissected formalin-fixed paraffin-embedded tumors from 693 patients with NSCLC were analyzed for EGFR mutations by allele-specific PCR in a pilot data set to test the hypothesis (pilot cohort). TTF-1 status was documented as positive, negative or not reported. Negative predictive value (NPV) for a range of true prevalences of EGFR mutation (1%-50%) was estimated using a Bayesian modeling approach. To further corroborate the hypothesis, a separate validation cohort of patients treated with erlotinib at two academic affiliated institutions with known TTF1 and EGFR mutation status was studied using the same modeling approach (validation cohort).

      Results
      301 patients with documented ACs and known TTF-1 status were included in the pilot cohort. In this population enriched to have EGFR mutations, EGFR mutations were present in 224 specimens (74%). Only 2 of the 224 specimens that were positive for EGFR mutations were negative for TTF-1 expression yielding a sensitivity of 99.1% (95% confidence interval (CI) 96.8-99.9%). For prevalence rates of EGFR mutations of 13% and 15%, the estimated NPV are 99.5% (95% credible interval (CRI) 98.6%-99.9%) and 99.4% (98.4%-99.9%), respectively. Data from 211 patients comprised the validation cohort. With an 11% rate of EGFR mutations, the estimated NPV was 92% (95% CRI - 73%-99%). For true EGFR mutation rates of 13% and 15%, using the data from the validation cohort, the estimated NPVs were 97% (95% CRI 92%-99%) and 96% (95% CRI 91%-99%), respectively. Figure 1. estimated NPVby true prevalence of EGFR mutation for both datasets Figure 1

      Conclusion
      An overwhelming majority of Lung ACs that are TTF-1 negative will be negative for EGFR mutations. These findings may be useful in avoiding delay of chemotherapy initiation in TTF-1 negative patients with newly diagnosed non-small cell lung cancer.

  • +

    P3.06 - Poster Session 3 - Prognostic and Predictive Biomarkers (ID 178)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Biology
    • Presentations: 1
    • +

      P3.06-051 - Circulating biomarkers may help guide selection of erlotinib versus chemotherapy in pretreated advanced NSCLC patients. (ID 3364)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      Recent data suggests a trend for prolonged progression-free survival (PFS) in patients with wild-type EGFR (wtEGFR) non-small cell lung cancer (NSCLC) treated with second-line docetaxel over erlotinib. In this same study, however, overall survival (OS) appeared unaffected. In the maintenance setting, erlotinib was also shown to improve both PFS and OS in wtEGFR advanced NSCLC patients. More recently, evaluating serum protein patterns by mass spectroscopy revealed inferior PFS with erlotinib compared to docetaxel in patients with a particular protein pattern. The individual proteins in the mass spectroscopy peaks were not identified. The objective of this study was to develop a multi-analyte serum panel of specific proteins with predictive value for erlotinib versus palliative chemotherapy in pretreated advanced NSCLC patients that were unselected for EGFR mutation status.

      Methods
      74 biomarkers were evaluated using Luminex immunobead assays against a total of 120 patients with stage IV NSCLC that were previously treated with chemotherapy and were unselected for EGFR mutation status. Patients were treated either with single agent erlotinib or platinum-based chemotherapy at the discretion of the treating physician. Patients were evaluated for disease progression using RECIST criteria and association of biomarker with survival outcomes was assessed using Cox proportional hazards regression model. The differential association of the biomarkers in the two treatment groups (erlotinib or chemotherapy) with survival outcomes was assessed using a proportional hazards (PH) interaction model.

      Results
      In univariate PH regression analysis, we identified 7 serum biomarkers (osteopontin, CA-125, sTNF-RII, PlGF, leptin, IGFBP5, and amphiregulin) which were strongly associated (p<0.01) and nine additional biomarkers (IGFBP4, sTNF-RI, CA 15-3, IGFBP1, sIL-2Rα, sFAS, VEGF-A, sIL-1RI, and sIL-4R) which had significant association (p<0.05) with PFS in the erlotinib subgroup (n=92). Similarly, 9 biomarkers (osteopontin, sTNF-RI, sTNF-RII, CA 15-3, sIL-2Rα, epiregulin, PILG, IL6 and CA 125) were found to be strongly associated with OS. In our assessment of differential association with PFS, we found eight serum biomarkers (sIL-2Rα, IL-8, sIL-1RI, Tensascin C, FGF2, HGF, Leptin, and TGF-α) with significant to strongly significant positive interaction terms, thus indicating differentially increased hazard of progression in the chemotherapy subgroup with high biomarker levels. Four markers (IL-8, TGF-α, HGF, VEGF-A) were found to have significantly positive interaction, indicating a decreased hazard of death in the erlotinib group with high biomarker levels; whereas two (sTNF-RII, PSA) had significant negative interaction with OS, demonstrating a increased hazard of death in the erlotinib group.

      Conclusion
      We identified a series of circulating surrogate biomarkers associated with epithelial-to-mesenchymal transition (EMT) that have promise for algorithm development to help physicians determine whether erlotinib as a single agent of chemotherapy is capable of improving outcomes in patients that progressed after first-line platinum-based chemotherapy. Current efforts focus on evaluating biomarker relationships with EGFR mutation status and algorithm validation in an effort to enhance survival in advanced stage NSCLC.

  • +

    P3.10 - Poster Session 3 - Chemotherapy (ID 210)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Medical Oncology
    • Presentations: 1
    • +

      P3.10-047 - Weight Gain During First-line Chemotherapy for Locally Advanced and Metastatic Non-Small-Cell Lung Cancer is Associated With Improved Survival (ID 2864)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      Two previous studies from our institution have shown that weight gain in the course of chemoradiotherapy for locally advanced or oligomestastatic NSCLC is associated with improved survival (Sher et al, 2013 and Gielda et al, 2010). This study aimed to determine the prognostic value of weight and serum albumin changes in patients with locally advanced and metastatic NSCLC receiving first-line platinum doublet chemotherapy.

      Methods
      Patients with newly-diagnosed locally advanced or metastatic NSCLC treated in the first-line setting with platinum doublet chemotherapy from June, 2011 to August, 2012 at RUMC were included for analysis. Weight and albumin values were recorded at baseline, 3, 6, and 12-week intervals from initiation of therapy. Their association with overall survival (OS) was assessed using Kaplan Meir methods and Cox proportional hazards regression.

      Results
      A total of 139 patients were included. Median age was 68 years (range 31-85). ECOG performance status (PS) was 0 in 29.5%, 1 in 46.7%, and 2 or greater in 16.5% of patients. Median baseline weight was 68.17 kg (range 40.1-123.4). Patients who experienced weight gain at 6 weeks had a significantly higher OS compared to those who lost weight, 20.4 months versus 13.6 months median survival respectively (log-rank p=0.025), Fig 1. In Cox regression analysis the hazard ratio (HR) for 1 kg of weight gain at 6 weeks was 0.84 (p <0.001). A marginal improvement in OS was seen for those who gained weight at 12 weeks, median survival not reached versus 15.5 months in those who lost weight (log-rank p=0.07). The HR for 1 kg weight gain by 12 weeks was 0.91 (p=0.002). Baseline albumin level was available for 97 patients. Median baseline albumin was 3.5 (range 1.5-4.7). A higher baseline albumin was found to be significantly associated with longer survival (HR 0.31, p<0.001). In a multivariate analysis, weight gain in 6 weeks was strongly associated (HR 0.81, p=0.003), and higher baseline albumin was associated (HR 0.49, p=0.03) with improved OS after adjusting for age, PS, and baseline weight. Figure 1: Figure 1

      Conclusion
      Higher baseline albumin and weight gain during first-line chemotherapy for locally advanced and metastatic NSCLC, appear to be associated with improved overall survival and may constitute important predictors of outcome. Study of molecular mechanisms involved in weight gain during anti-neoplastic treatment might provide ideas for novel therapeutic strategies.

  • +

    P3.11 - Poster Session 3 - NSCLC Novel Therapies (ID 211)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Medical Oncology
    • Presentations: 1
    • +

      P3.11-039 - Exploration of patient health status as measured by the generic preference-based questionnaire EQ-5D alongside the START trial of tecemotide (L-BLP25) in non-small cell lung cancer (ID 2744)

      09:30 - 16:30  |  Author(s): P. Bonomi

      • Abstract

      Background
      Tecemotide (L-BLP25) is a mucin 1 (MUC1) antigen-specific cancer immunotherapy investigated in patients not progressing after primary chemo-radiotherapy for stage III non-small cell lung cancer (NSCLC) in the phase III START study. The objective of this analysis was to explore patients’ health status alongside the study.

      Methods
      From January 2007 to November 2011, 1513 patients with unresectable stage III NSCLC that did not progress after chemo-radiotherapy (platinum-based chemotherapy and ≥50 Gy) were randomized (2:1; double-blind) to tecemotide (806 μg lipopeptide) or placebo SC weekly x 8 then Q6 weeks until disease progression or withdrawal. The analysis population (n=1239) was defined prospectively to account for a clinical hold of the study. The impact on patient health status was assessed as an exploratory endpoint using the EuroQoL 5 Dimensions (EQ-5D), a widely used generic preference-based questionnaire covering 5 dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression). EQ-5D index score can be calculated for which perfect health is given a value of 1 and death a value of 0. EQ-5D was collected at baseline, weeks 2, 5 and 8 and then every 6 weeks until end of treatment (EOT) visit (i.e. at time of disease progression), the EOT visit and every 12 weeks afterwards. Analysis of covariance (ANCOVA) was carried out to explore the change of EQ-5D index score over time in the overall population for patients on treatment. The change of EQ-5D to EOT visit was also estimated. Change of EQ-5D index score was explored using all data (i.e. collected both before and after EOT visit) using a linear growth curve model, with random intercept and slope, considering time as a continuous variable.

      Results
      EQ-5D compliance rates (percentage of patients still in the study who completed the questionnaire) were consistently above 85% for all visits of the treatment period in both treatment arms. Mean baseline EQ-5D score was 0.79 (sd=0.19) for both tecemotide and placebo arms. The results from ANCOVA on the overall population did not show any significant difference between the two arms during the treatment phase. Change in the EQ-5D index score from baseline to EOT visit was –0.102 (95%CI: –0.134, –0.071) for tecemotide and –0.136 (95%CI: –0.177, –0.095) for placebo. The linear growth model including the EQ-5D assessments before and after EOT showed that the EQ-5D index score decreased significantly over time in both treatment arms, but that the decrease was slightly slower in the tecemotide than in the placebo arm: –0.0076 per month in tecemotide patients vs. –0.01 in placebo (p=0.0498).

      Conclusion
      During treatment, there was no statistical difference in health status with tecemotide vs. placebo. This supports the good tolerability profile of tecemotide, with a lack of significant toxicity as compared to placebo. Disease progression was associated with a notable deterioration of patient health status, regardless of the treatment. Considering data from both before and after disease progression, patients’ health status appeared to worsen slightly over time, at a slower rate for patients treated with tecemotide.