Virtual Library

Start Your Search

Melania Pintilie



Author of

  • +

    MA25 - Oligometastasis: Defining, Treating, and Evaluating (ID 929)

    • Event: WCLC 2018
    • Type: Mini Oral Abstract Session
    • Track: Oligometastatic NSCLC
    • Presentations: 1
    • Now Available
    • Moderators:
    • Coordinates: 9/26/2018, 13:30 - 15:00, Room 203 BD
    • +

      MA25.11 - Clinical and Molecular Predictors of Outcome in Patients with EGFR mutant NSCLC Brain Metastases treated with RT (Now Available) (ID 14529)

      14:40 - 14:45  |  Author(s): Melania Pintilie

      • Abstract
      • Presentation
      • Slides

      Background

      Brain metastases(BM) develop in ~45% of patients with EGFR mutant(EGFRm) non-small cell lung cancers(NSCLC). There are limited reports on clinical/molecular factors associated with BM outcomes after radiotherapy in EGFRm NSCLC patients.

      Method

      We identified patients with EGFRm NSCLC who presented with or developed BM and had their lung tumor resected. Clinical, demographic and TP53 status were collected from medical/pathology records. Whole-Exome Sequencing of the primary tumor was performed. Overall survival(OS) and intracranial progression(IP) were defined from start of BM treatment and correlated with clinical/molecular features. IP was defined from the date of BM treatment until any brain failure, either local(previously present BM) or distant(development of new BM). Categorical and continuous covariates were tested by Fisher exact or Mann-Whitney test, respectively. OS by Kaplan-Meier with groups compared by log-rank. For each model the Harrell Concordance Index(CI) was performed.

      Result

      From 41 eligible patients with BM, 9 were excluded due to sequencing quality. Of the 32 remaining patients, 20 (62%) had their BM treated with WBI (15 WBI alone and 5 TKIàWBI), 12 (38%) with TKI±SRS (9 TKI àSRS; 2 TKI alone and 1 SRS alone). Median age at BM was 59.5 years(y). Most of the cohort were female(81%), non-smoker(78%), non-Asian(62%) and 50% presented as stage III or higher at diagnosis. An EGFR exon 19 mutation was present in 72% of patients, 25% had 2 or more EGFRm, 15% with additional driver mutations and 53% with TP53 co-mutation. At a median follow-up of 1.21-y, no clinical/molecular factors(treatment, age, gender, ethnicity, smoking status, stage at presentation, EFGR exon 19 versus 21, number of EGFRm, additional driver mutations, TP53 co-mutation) correlated with survival. There was a trend for longer survival for patients treated with TKI±SRS(median 3.4y) compared to WBRT±TKI(median 1.4y); p=0.08 and for age at BM ≤59.5y(median 2.5y) compared to >59.5y (median 1.4y); p=0.2. Higher risk of IP was observed in younger patients (age as continuous variable) with HR of 0.94(95%CI 0.88-1.0), p=0.04; favoring older patients and remained significant after accounting for treatment modality on multivariate analysis p=0.03. No additional clinical/molecular factors correlated with IP.

      Conclusion

      In our study, younger age at BM treatment was associated with higher IP. We also observed a trend for longer OS for younger patients(≤59.5y) and for patients treated with TKI±SRS. Our data suggest that younger patients with EGFR BM should undergo close intracranial follow up and that future studies to define the benefit of brain-directed multimodality treatment are warranted.

      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.

  • +

    MTE01 - Preclinical Models of Lung Cancer (Ticketed Session) (ID 811)

    • Event: WCLC 2018
    • Type: Meet the Expert Session
    • Track: Biology
    • Presentations: 1
    • Now Available
    • Moderators:
    • Coordinates: 9/24/2018, 07:00 - 08:00, Room 206 F
    • +

      MTE01.02 - Lung Patient Derived Xenograft and Organoid (Now Available) (ID 11547)

      07:30 - 08:00  |  Author(s): Melania Pintilie

      • Abstract
      • Presentation
      • Slides

      Abstract

      Establishment of preclinical lung cancer models that closely match patient tumor biology is imperative for developing therapeutic strategies with the most translational relevance. Non small cell lung cancer (NSCLC) cell lines grown under 2D conditions or as cell line-derived xenografts (CDXs) are the most widely used models. They have been complemented with murine models engineered to develop lung cancer after introduction of specific genetic alterations (GEMMs). Lung cancer cell lines are readily amenable to mechanistic studies and economical high-throughput drug screening. However, for many NSCLC cell lines, the spectrum of mutations and copy number alterations have drifted considerably relative to patient tumors 1. This finding, in conjunction with long-term adaption to heterologous in vitro growth conditions, raise concerns about the extent to which cell line biology and potential drug responses may have deviated from clinical tumors. GEMMs are powerful tools for studying specific oncogenic mechanisms in isolation in vivo, but these models lack the intratumoral heterogeneity of patient tumors, which is thought to play a major role in the development of drug resistance. Furthermore, ideally, GEMMs should be constructed using the appropriate cell of origin context, which is challenging, as there are differences in the compositions of human and murine airways, and the cellular origins for most forms of lung cancer have not been established.

      NSCLC patient-derived xenografts (PDXs) overcome some of the limitations of these other models. They show much less genetic drift than cell lines, and their mRNA expression and the phospho-tyrosyl proteome more closely match patient tumors 1, 2. We have established a large collection of NSCLC PDXs from surgically resected tumors and endobronchial ultrasound-guided (EBUS) and CT-guided biopsies. Tumor specimens were initially implanted in the subcutaneous flanks of NSG mice (NOD SCID gamma, non-obese diabetic severe combined immunodeficiency, gamma). The PDX tumors have been viably cryopreserved and can be serially passaged in NOD SCID mice. Most of our collection comprises the major histologic subtypes of NSCLC [52 adenocarcinomas (LUAD) and 62 squamous cell carcinomas (LUSC)]. They, along with the primary patient tumors, are being molecularly profiled at multiple levels so that they can be optimally used for personalized medicine studies and novel integrated approaches to understand NSCLC pathogenesis, prognosis, and treatment. These levels include copy number variations, exome mutations, DNA methylation, mRNA and miRNA expression, and proteomics. In general, the PDX models recapitulate the mutation spectrum, copy number variations, and gene expression of matched patient histologies. They also recapitulate sensitivity and resistance to known targeted therapeutics (e.g. EGFR inhibitors), and thus, can be used to dissect mechanisms underlying differential drug responses. Such studies are ongoing, including investigation of potentially new biomarker-targeted therapeutic combinations. We have also found that not all patient tumor fragments engraft successfully, and that successful engraftment correlates with poor prognosis of the patient 3. We are using this relationship to discover a new molecular fingerprint to predict clinical outcome, as well as understand the bases that distinguish less and more aggressive tumor behavior.

      In parallel, we have developed methods to grow organoids from primary patient tumors and PDX models in 3D culture using Matrigel (PDO and XDO, respectively). For LUAD, both the PDO and XDO success rate of establishing bona fide organoid models is ~20%. Our stringent criteria include a minimum capacity of 10 passages and a split ratio of at least 1:3. LUSC has been more difficult to establish as organoid models, with a success rate of 17%, and only from PDXs, so far. Using these methods, we have established 4 models of each histology, which we have confirmed form tumors when transplanted into mice. Molecular profiling indicates that the organoids maintain the same mutation spectrum and copy number variations of their parental tumor tissue. These models offer distinct advantages over PDXs and cell lines. As compared to standard 2D cultures, they recapitulate the appropriate tissue histology, and thus, possibly clinically relevant growth control mechanisms, even while growing ex vivo. This notion is further supported by the ex vivo conditions supporting gene expression patterns, which allows the organoids to be segregated into their respective tumor histologies when using signatures derived from patient or PDX material. Given the low cost, rapid growth rates, and ease of in vitro manipulation, these models are ideally suited for rapid discovery and testing of new therapeutic strategies that can be matched to specific patient molecular profiles.

      In summary, generation of molecularly profiled PDX and organoid models offer great opportunity for translational and personalized medicine in NSCLC.

      1. Gao, H. et al. High-throughput screening using patient-derived tumor xenografts to predict clinical trial drug response. Nat. Med. 21, 1318-1325 (2015).

      2. Wang, D. et al. Molecular heterogeneity of non-small cell lung carcinoma patient-derived xenografts closely reflect their primary tumors. Int. J. Cancer 140, 662-673 (2017).

      3. John, T. et al. The ability to form primary tumor xenografts is predictive of increased risk of disease recurrence in early-stage non-small cell lung cancer. Clin. Cancer Res. 17, 134-141 (2011).

      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.

  • +

    P2.11 - Screening and Early Detection (Not CME Accredited Session) (ID 960)

    • Event: WCLC 2018
    • Type: Poster Viewing in the Exhibit Hall
    • Track:
    • Presentations: 1
    • Moderators:
    • Coordinates: 9/25/2018, 16:45 - 18:00, Exhibit Hall
    • +

      P2.11-19 - MicroRNAs as Liquid Biopsy Biomarkers for Early Detection in Lung Cancer. (ID 14050)

      16:45 - 18:00  |  Author(s): Melania Pintilie

      • Abstract
      • Slides

      Background

      microRNAs (miRNAs) control the expression of key driver genes associated with tumorigenesis in several cancer types and can be detected as stable circulating molecules in body fluids. Deregulated miRNAs have been identified as potential biomarkers in plasma from cancer patients. In lung cancer, the most promising clinical application of circulating miRNAs is early disease detection, since late diagnosis is a major clinical problem associated with patient death.

      Method

      38 plasma samples from patients with lung adenocarcinoma and squamous cell carcinoma and 21 healthy controls from a screening population were profiled for an 800 miRNA set using the Nanostring nCounter® platform. Validation was performed in an independent sample set of 40 patients and 40 controls, paired by age and sex, using TaqMan® quantitative real-time PCR. Statistical analyses were performed using the Mann Whitney test, and miRNA signatures identified by Elastic net, improved Maximizing R Square Analysis (MARSA) and C-Statistics. Bioinformatic approaches were applied for external data validation against other miRNA expression datasets.

      Result

      A subset of 149 miRNAs was significantly over-expressed in patient plasma compared to controls with fold change ≥2, p≤0.01 and FDR<0.05. In addition, three distinct miRNA signatures with 12 unique miRNAs were identified in the discovery set (hsa-miR-16-5p, hsa-miR-92a, hsa-miR-106b-5p, hsa-miR-148b-3p, hsa-miR-155-5p, hsa-miR-217, hsa-miR-378e, hsa-miR-451, hsa-miR-484, hsa-miR-1285-3p, hsa-miR-1285-5p and hsa-miR-664a-3p). All signatures were validated in the independent sample set, being able to distinguish patients from controls. Interestingly, miRNAs identified herein control the expression of tyrosine kinase, transcription factors and immune system related genes associated with lung tumorigenesis.

      Conclusion

      By using a highly specific and sensitive assay and stringent criteria on sample selection and data analyses, we were able to identify known and novel miRNAs that are significantly deregulated exclusively in plasma from patients with a diagnosis of lung adenocarcinoma or squamous cell carcinoma. Our results contribute to the identification of circulating plasma miRNAs as potential biomarkers for early disease detection in lung cancer.

      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.

  • +

    P3.03 - Biology (Not CME Accredited Session) (ID 969)

    • Event: WCLC 2018
    • Type: Poster Viewing in the Exhibit Hall
    • Track:
    • Presentations: 2
    • Now Available
    • Moderators:
    • Coordinates: 9/26/2018, 12:00 - 13:30, Exhibit Hall
    • +

      P3.03-18 - Collagen Type XI Promotes Lung Adenocarcinoma Dissemination Via Integrin α2 and DDR1 (ID 14009)

      12:00 - 13:30  |  Author(s): Melania Pintilie

      • Abstract

      Background

      Our previous studies showed that non-small cell lung cancer (NSCLC) stroma is an important actor of lung tumorigenesis. We found integrin α11, a fibrillar collagen receptor expressed on cancer-associated fibroblasts (CAF), to be strongly up-regulated in NSCLC stroma and to promote NSCLC growth in vivo. Collagen XI is a fibrillar collagen that expression is increased in NSCLC and is correlated with integrin α11 expression. The purpose of the present study is to determine whether integrin α11-mediated lung tumorigenesis is induced via collagen XI and how that collagen may affect lung carcinoma cell invasion.

      Method

      To study the effects of collagen XI, collagen XI has been incorporated in a matrix of collagen I at a ratio 1:3 (collagen I/XI). For tumor cell invasion, lung adenocarcinoma H1975 cells were allowed to form spheroids before being embedded in a collagen I or a mix of collagen I/XI matrix. Area of dissemination and migration pattern has been analyzed 48 hours later. In order to identify the collagen receptor interacting with collagen XI, we used cell attachment and ELISA assay. Knockout of integrin α2 and DDR1 has been performed using the CRISPR-Cas9 strategy in H1975 cell line.

      Result

      We showed that high collagen XI expression is associated with high risk of recurrence in the UHN cohort of 165 NSCLC patients (hazard ratio= 1.97, 95% confidence interval 1.09-3.56, P<0.026). Surprisingly, CAF embedded in a mix of collagen I/XI matrix failed to reorganize the collagen matrix as efficient as embedded CAFs in only collagen I matrix and displayed inhibition of migration in presence of collagen XI. In contrast, collagen XI greatly enhanced invasion of lung adenocarcinoma H1975 cells in a mix collagen matrix with a different invasion pattern compared to H1975 cell invasion in collagen I. Although integrin α11 expression correlates with collagen XI expression in NSCLC, this integrin bound poorly to collagen XI. Instead we found integrin α2 and DDR1 to interact with collagen type XI (Kd=5nM and Kd=24nM, respectively). Knockout of integrin α2 or DDR1 in H1975 cells suppressed collagen XI-increased invasion in a collagen I/XI matrix, indicating that both collagen receptors are required to mediate collagen XI effect on lung adenocarcinoma dissemination.

      Conclusion

      We showed that the presence of collagen XI in NSCLC stroma could enhance lung adenocarcinoma dissemination via integrin α2 and DDR1. Thus, targeting these two collagen receptors may be a new strategy to improve the outcome of lung cancer.

    • +

      P3.03-29 - The Prognostic Effect of Tumor Mutation Burden and Smoking History in Resected EGFR Mutant Non-Small Cell Lung Cancer (Now Available) (ID 13199)

      12:00 - 13:30  |  Author(s): Melania Pintilie

      • Abstract
      • Slides

      Background

      Although EGFRm NSCLC occurs mainly in non-smoking patients, most series report 20%-35% of cases in current or previous smokers. Broad molecular profiling of EGFRm NSCLC in smokers has not been reported.

      Method

      Surgically resected primary EGFR exon 19, 20 and 21 mutated NSCLC tumors from 106 patients were molecularly profiled by whole exome sequencing using the Illumina HiSeq2000 platform. Alignment and variant discovery analysis was performed according to GATK best practices workflow; 74 sequenced to a mean coverage of 65.1x. Demographics and outcomes were compared for smokers and non-smokers (non-S), and by mutation profile.

      Result

      Among 53 non-smokers and 21 smokers (5 current/recent within 10 years), 70% were female, 51% non-Asian, 40.5% >65 years of age and 58.1% had EGFR exon 19. Of the 74 patients, 51% were stage I, 19% stage II and 30% stage III+. Smoking was associated with male sex (p= 0.011) and non-Asian ethnicity (p=0.00002) but not age, stage or EGFR exon 19/20/21 subtype. Multiple “driver” mutations occurred in tumors of 23.8% smokers and 26.4% non-S. TP53/EGFR co-mutation occurred in 52.4% smokers and 47.2% non-S. There was no significant difference seen for TMB in smokers: median TMB in smokers 3(1.4-7.46) versus 2.7(0.96-3.95) in non-S (p=0.11). The strongest prognostic factor for OS and DFS was stage (I, II, III+) (p<0.0001 for each compared to stage I). Smoking history did not have a significant effect on survival: HR 1.61 (CI 0.78-3.32, p=0.2) or probability of relapse: HR 0.9 (CI 0.46-1.77, p=0.77). Smoking within 10 years of NSCLC diagnosis was not associated with shorter OS/DFS. Neither EGFRm subtype nor TP53/EGFR co-mutation was associated with probability of relapse or OS. High TMB was significantly associated with shorter survival: HR above vs below the median 2.37 (CI 1.12-5.01, p=0.024), but not probability of relapse: HR 1.67 (CI 0.88-3.19, p=0.12). A subset analysis found the effect of TMB on survival was significant in patients >65 years (p=0.004), but not in patients <65 years (p=0.95).

      Conclusion

      Stage remains the strongest prognostic factor for survival and probability of relapse in completely resected EGFRm NSCLC. TMB appears to have an effect on survival outcomes, unlike smoking status, and this effect may be greater in patients older than 65 years.

      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.

  • +

    P3.16 - Treatment of Early Stage/Localized Disease (Not CME Accredited Session) (ID 982)

    • Event: WCLC 2018
    • Type: Poster Viewing in the Exhibit Hall
    • Track:
    • Presentations: 1
    • Now Available
    • Moderators:
    • Coordinates: 9/26/2018, 12:00 - 13:30, Exhibit Hall
    • +

      P3.16-07 - The Impact of Clinical and Molecular Profile of Resected EGFR-Mutant Non-Small Cell Lung Cancer on the Risk of Developing Brain Metastases (Now Available) (ID 13339)

      12:00 - 13:30  |  Author(s): Melania Pintilie

      • Abstract
      • Slides

      Background

      Brain metastases are common in non-small cell lung cancers (NSCLC) with activating EGFR mutations (EGFRm), occurring in 44%-63% of patients. To date, there are no known clinical or molecular factors to predict the risk of brain metastases in these patients.

      Method

      In this retrospective single-institution study, we identified 106 patients with EGFRm NSCLC who underwent surgery for primary lung tumor. Clinical and demographic data was collected from electronic records. Whole Exome Sequencing (WES) of the primary tumor was performed utilizing the Agilent SureSelect Exome v6+COSMIC baits followed by sequencing on the Illumina HiSeq2500 platform. Development of brain metastases was correlated with clinical/pathologic features, EGFR mutation type, co-mutation of EGFR and other frequently mutated genes; and non-synonymous tumor mutation burden (TMB). Statistical analysis used Fisher exact test for categorical variables, Mann-Whitney test for continuous variables of association with the risk of developing brain metastases, and Gray’s test for the probability of brain metastases over time.

      Result

      Of 106 patients who underwent surgical resection of primary EGFRm NSCLC, WES was successful for 73: 51 (70%) females, 52 (71%) never smokers, 38 (52%) stage I, 14 (19%) stage II and 21 (28%) stage III; 42 (57%) EGFR exon 19 mutation, 30 (41%) exon 21, 1(1%) Exon 20 insertion mutation.

      Twenty-five patients (34%) developed brain metastases. Patients with brain metastases were younger (median age 61 vs. 65 years, p=0.021), had more advanced stages (p=0.012), with a trend towards higher rates in females (p=0.066). One patient with brain metastases had de-novo EGFR T790M mutation in the primary tumor. No difference was seen regarding smoking history, EGFR mutation type, TP53 co-mutation, and median TMB. The 5-year probability of brain metastases increased with increasing stage (14% stage I; 43% stage II, [HR=3.00], 44% stage III, [HR=3.13], p=0.03), and a trend towards higher probability among females (33% vs. 19%; HR=0.39 for males, p=0.074), and younger patients (37% <65 years vs. 15% >65, HR=0.37 in older patients, p=0.042). There was no difference in probability of brain metastases based on smoking history, ethnicity, EGFR type (33% exon 19 vs. 22% exon 21, p=0.28), TP53 co-mutation (31% vs. 27% without TP53, p=0.59), or TMB (24% TMB≤2.87 vs. 32% TMB>2.87non-synonymous mutations/Mb, p>0.99).

      Conclusion

      While our findings suggest that younger age, advanced stage, and female sex may be associated with the development of BM in EGFRm NSCLC, we could identify no molecular predictor of BM based on EGFR subtype, TP53 co-mutation or TMB.

      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.