Virtual Library

Start Your Search

Ming Sound Tsao

Moderator of

  • +

    GR03 - Problem Areas for the Next WHO Classification of Lung Cancers (ID 31)

    • Event: WCLC 2019
    • Type: Grand Rounds Session
    • Track: Pathology
    • Presentations: 5
    • Now Available
    • +

      GR03.01 - High Grade Neuroendocrine Tumors (Now Available) (ID 3309)

      15:45 - 17:15  |  Presenting Author(s): MAURO GIULIO PAPOTTI

      • Abstract
      • Presentation
      • Slides

      Abstract

      The four-tier WHO 2015 classification scheme of lung neuroendocrine neoplasms (NEN) includes morphologically and clinically heterogeneous conditions (1). High grade tumors typically encompass large and small cell neuroendocrine carcinomas (LCNEC and SCLC, respectively). The survival of these two types of poorly differentiated NENs is similar and significantly different from that of the well differentiated carcinoid tumors. Thus their appropriate classification is a clinically relevant exercise. While the morphological features of classical SCLC and those of low grade carcinoids allow to easily take these two tumors apart, the correct classification of some atypical carcinoids and of LCNEC is less straightforward.

      In fact, morphology alone may not be sufficient to identify the various histotypes (that still represent the most relevant prognostic parameter in NENs), and to specifically classify aggressive forms into the group of high grade carcinomas. Even immunophenotype profiling may fail to some extent and only the more recent genetic data have been able to better stratify variations within each single histological type (as recognized by the current WHO criteria). According to such criteria (1), high grade neuroendocrine (NE) carcinomas are defined as malignant tumors made of large or small cells having a solid, diffuse (or more rarely irregularly organoid) growth patterns, with extensive necrosis and a mitotic index exceeding 10 per 10 high power fields. This definition fits for a relatively wide group of tumors, whose clinical behavior is not perfectly overlapping. In particular, while SCLC are invariably associated to a high mitotic rate and high grade cytological features including classical salt&pepper chromatin pattern, LCNEC belong to a grey area that merges with atypical carcinoids on the one side (having intermediate values of mitotic index), and with SCLC on the other (with the occurrence of combined small and large cell NE carcinoma variants).

      Immunophenotypic markers are not always useful for accurately stratifying NENs. In fact, chromogranin A, synaptophysin and CD56 are generally expressed by the majority of NENs, though with a different intensity and distribution (for example, SCLC may be negative or only focally reactive for chromogranin, but invariably expresses synaptophysin) (2,3). Some transcription factors such as TTF1 and hASH1 are usually intensely positive in high grade tumors, both LCNEC and SCLC, as opposed to carcinoid tumors, that are generally not reactive (with the possible exception of some peripherally located spindle cell carcinoids).

      The proliferation index, as defined by Ki-67 immunohistochemistry, was proposed as an effective complementary tool to identify different prognostic subgroups, although its use is not officially accepted by the WHO classification with the exception of a differential diagnostic role in small biopsy specimens (1). Indeed, high grade tumors have a much higher mean Ki67 index compared to carcinoids (mean values of 60-80% versus 2-8%). For this reason, the integration of Ki67 data with the two official morphological parameters (necrosis and mitoses) proved effective in a proposed grading system (4).

      The spectrum of aggressive NENs is unfortunately complicated by the existence of combined NENs, having areas of brisk proliferation admixed with a relatively quiescent tumor cell population. In addition, rare cases have been demonstrated to progress from well differentiated carcinoid to high grade NE carcinomas. The relationship between low and high grade NENs is further supported by the observed heterogeneous genetic profile of high grade tumors, namely LCNEC. Apart from the original detection of the carcinoid-specific MEN1 mutations in a small fraction of “morphological” LCNECs, and of two other different groups of LCNEC, one related to SCLC and the other associated to a genetic signature of non small cell lung carcinomas (5,6), recent comprehensive genomic and transcriptomic analyses of 75 LCNEC identified two molecular subgroups, labeled "type I LCNEC" (having bi-allelic TP53 and STK11/KEAP1 gene alterations, and a NE profile with ASCL1 high / DLL3 high / NOTCH low), and "type II LCNEC" (enriched for bi-allelic inactivation of TP53 and RB1 genes, reduced NE markers, ASCL1 low / DLL3 low / NOTCH high, upregulation of immune-related pathways) (7). In this latter study, some genomic alterations were shared with pulmonary adenocarcinomas and squamous cell carcinomas.

      In a more recent study (8), the reverse approach was used, starting from a series of carcinoid tumors. With the aim of a full molecular NEN characterization by integrative analyses of genomic, transcriptomic, and methylome data, three molecular groups were identified: clusters A through C were enriched by typical carcinoids (TC), atypical carcinoids (AC) and LCNEC, respectively. Interestingly, the latter cluster also included a subgroup of six “morphological” ACs, here designated “supra-AC” that were molecularly similar to LCNEC, thus supporting the postulated link between the low and high grade lung NENs. Therefore, misclassification is common between AC and LCNEC, due to the existence of “carcinoid-like” LCNEC (5,7), possibly resulting from the evolution of a well- into a poorly differentiated NEN (9), as also reported in thymic LCNEC (10).

      In conclusion, the correct classification of high grade lung NENs is in general easily obtained in conventional forms of oat cell SCLC and of highly atypical and proliferating LCNEC. Conversely, the separation is more subtle in the presence of the rare intermediate (grey zone) cases, standing between AC and LCNEC, that probably correspond to the recently proposed category of “G3 NE Tumor” in the pancreas (11) and the gastrointestinal tract (expected in the next WHO classification of digestive system NENs).

      References

      1 Travis et al. WHO classification of tumors of the lung. IARC press, Lyon, 2015

      2 Thunissen E et al. J Thorac Oncol 2017;12:334-346

      3 Yatabe et al. J Thorac Oncol 2019;14:377-407

      4 Rindi G et al. Endocr Rel Cancer 2013;21:1-16

      5 Rekthman N et al. Clin Cancer Res 2016; 22, 3618-3629

      6 Simbolo M et al. J Pathol 2017; 241: 488–500

      7 George J et al. Nat Commun 2018; 9: 1048 (1-13)

      8 Alcala N et al. Nat Commun 2019 (in press)

      9 Pelosi G et al. Virchows Arch 2018;472:567-577

      10 Fabbri A et al. Virchows Arch 471, 31-47

      11 Lloyd RV et al. WHO classification of endocrine tumors. IARC press, Lyon, 2017

      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.

    • +

      GR03.02 - Adenocarcinoma (Now Available) (ID 3310)

      15:45 - 17:15  |  Presenting Author(s): Andrew G Nicholson

      • Abstract
      • Presentation
      • Slides

      Abstract

      Many of the limitations in the WHO 2004 classification for adenocarcinomas (confusion over the term bronchioalveolar adenocarcinoma, usage of the term “mixed pattern” and no classification for small biopsy/cytology specimens) were addressed in the 2011 IASLC/ATS/ERS multidisciplinary classification,1 and this proposal was adopted by the 2015 WHO classification with minor changes.2

      Adenocarcinoma in situ, minimally invasive adenocarcinoma, lepidic predominant adenocarcinoma and invasive mucinous adenocarcinoma have now replaced the term “bronchioloalveolar adenocarcinoma”, with subsequent TNM staging changes to use invasive size for T factor size.3 However there remains significant interobserver variation between pathologists in relation to the point at which invasion starts, indicating that tighter definitions of distinguishing a lepidic pattern from other invasive patterns are needed, especially in areas where these disagreements have a clinical impact.

      Interobserver agreement between pathologists has been shown to be much better for distinction of invasive patterns (acinar, papillary, micropapillary, solid). Accumulating data supports the 2015 WHO proposal that the cribriform pattern be regarded as a pattern with adverse prognostic significance. It is also proposed that micropapillary be expanded to include a filigree, as well as classical, pattern. One of the most important needs is for pathologists to better recognize the morphologic spectrum of the micropapillary pattern which is often underestimated. Several publications suggest prognostic groupings as lepidic, acinar/papillary and solid/micropapillary as a stratification, and these have been shown to predict response to adjuvant therapy.4 This leads into the issue of grading of resected adenocarcinomas and the presence of more aggressive histological patterns as a minor component.

      The histological feature termed “spread through airspaces” or STAS has been shown to be a poor prognostic factor for all major histologic types of lung cancer, including adenocarcinoma where it is frequently seen. There is considerable evidence that the presence of STAS carries prognostic significance,5,6 in particular in relation to non-anatomic resections, but there remains a need to identify where STAS begins and artefactual dissemination of tumour due to handling and processing of specimens ends.7 A tighter definition and evidence of international reproducibility is needed.

      While subtyping of histological patterns is well established in non-mucinous adenocarcinomas, mucinous adenocarcinomas are less well characterised. Various patterns of mucinous differentiation have been proposed, as well as assignment of histologic patterns in similar fashion to non-mucinous ADCs although only invasive mucinous adenocarcinomas (IMA) and colloid adenocarcinomas currently have specific subgroupings.2 This proposal has proved to be well founded given the specific molecular features and behaviour pattern of IMAs,8 although again work is required to refine prognostication. More data is also required tumours with mixed mucinous and non-mucinous areas.

      Resections are increasingly occurring after neoadjuvant therapy, with there is already a need to assess these in a structured fashion.9,10 Work is ongoing within the IASLC Pathology Committee to propose a method for classification in this clinical scenario.

      The 2015 WHO classification saw a seminal change in its structure, in that a classification system was proposed for biopsies and cytology specimens, rather than solely resections. In addition, a major theme utilized in the 2015 WHO classification was a multidisciplinary approach incorporating surgery, imaging, oncologic respiratory medicine, molecular biology as well as pathology. which needs to be maintained into the discussions of future classifications.1 This approach must remain and will likely need to be enhanced, given the revolution in molecular and immunologic characterisation of tumours, especially adenocarcinomas, and all these new clinically relevant findings will need to be part of pathologic reporting for the ensuing decades. The relative importance and structure of morphologic, immunohistochemical, molecular and immunologic data will need to be incorporated into a system that is appropriate not just for the most advanced cancer centres where all data are available but for laboratories and diagnostic services in underserved countries where morphologic features may be the only ones available.

      REFERENCES

      1. Travis WD, Brambilla E, Noguchi M, et al. International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society: international multidisciplinary classification of lung adenocarcinoma: executive summary. Proc Am Thorac Soc 2011;8:381-5.

      2. WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart. Lyons, France.: International Agency for Research on Cancer (IARC); 2015.

      3. Travis WD, Asamura H, Bankier AA, et al. The IASLC Lung Cancer Staging Project: Proposals for Coding T Categories for Subsolid Nodules and Assessment of Tumor Size in Part-Solid Tumors in the Forthcoming Eighth Edition of the TNM Classification of Lung Cancer. J Thorac Oncol 2016;11:1204-23.

      4. Tsao MS, Marguet S, Le Teuff G, et al. Subtype Classification of Lung Adenocarcinoma Predicts Benefit From Adjuvant Chemotherapy in Patients Undergoing Complete Resection. J Clin Oncol 2015;33:3439-46.

      5. Chen D, Mao Y, Wen J, et al. Tumor Spread Through Air Spaces in Non-Small Cell Lung Cancer: a systematic review and meta-analysis. Ann Thorac Surg 2019.

      6. Kadota K, Nitadori J, Sima CS, et al. Tumor Spread through Air Spaces is an Important Pattern of Invasion and Impacts the Frequency and Location of Recurrences after Limited Resection for Small Stage I Lung Adenocarcinomas. J Thorac Oncol 2015;10:806-14.

      7. Blaauwgeers H, Flieder D, Warth A, et al. A Prospective Study of Loose Tissue Fragments in Non-Small Cell Lung Cancer Resection Specimens: An Alternative View to "Spread Through Air Spaces". Am J Surg Pathol 2017;41:1226-30.

      8. Fernandez-Cuesta L, Plenker D, Osada H, et al. CD74-NRG1 fusions in lung adenocarcinoma. Cancer Discov 2014;4:415-22.

      9. Qu Y, Emoto K, Eguchi T, et al. Pathologic Assessment After Neoadjuvant Chemotherapy for NSCLC: Importance and Implications of Distinguishing Adenocarcinoma From Squamous Cell Carcinoma. J Thorac Oncol 2019;14:482-93.

      10. Blumenthal GM, Bunn PA, Jr., Chaft JE, et al. Current Status and Future Perspectives on Neoadjuvant Therapy in Lung Cancer. J Thorac Oncol 2018;13:1818-31.

      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.

    • +

      GR03.03 - Pleomorphic Carcinomas (Now Available) (ID 3311)

      15:45 - 17:15  |  Presenting Author(s): Alain Borczuk

      • Abstract
      • Presentation
      • Slides

      Abstract

      The category of sarcomatoid carcinoma in lung cancer classification is composed of five tumor types - pleomorphic, spindle, giant cell, carcinosarcoma and blastoma. While these are all relatively rare tumors, the pleomorphic carcinoma category is the most common of this group. Pleomorphic carcinomas are defined as combinations of adenocarcinoma, squamous carcinoma or large cell carcinoma with a spindle or giant cell element. It may be that spindle or giant cell examples, while diagnostically more challenging, represent variants of similar histogenesis but with complete mesenchymal transformation. Small cell carcinoma in a pleomorphic carcinoma is exceedingly rare. These tumors are often bulky tumors at presentation, with a propensity for central necrosis. Historically, this tumor is highly aggressive and treatment refractory.

      The histology of this tumor type includes correct identification of a malignant spindle component morphologically, or a giant cell component. While nuclear pleomorphism is an aspect of the tumor, the degree of nuclear enlargement, multinucleation and the presence of emperipolesis all distinguish giant cells of pleomorphic carcinoma from nuclear enlargement in high grade tumors. Immunohistochemistry has be helpful in identifiable a cytokeratin positive spindle or giant cell component. The use of zinc finger E-box binding homeobox1 (ZEB1), a protein involved in epithelial-mesenchymal transition to identify spindle or giant cell component of these tumors, both in small samples and resections, is emerging.

      Molecular alterations have also been linked to pleomorphic carcinomas. The tumors harbor mutations in KRAS as well as a higher rate of MET exon 14 skipping mutations. This is generally confined to cases with an adenocarcinoma component. TP53 mutations are also frequent. The molecular mechanisms of pleomorphic carcinoma with a squamous only epithelial component remain to be characterized.

      It has been proposed that MET exon 14 mutations may be targetable using agents such as crizotinib. In addition, these tumors show an elevated rate of high positive PDL1 immunoreactivity which may offer immunotherapy option in these patients.

      The relationship between large cell carcinoma and new entities such as SMARCA4 deficient carcinoma/sarcoma and the category of sarcomatoid carcinoma remains unclear. Greater elucidation of the molecular underpinning of sarcomatoid carcinoma categories may help clarify the place for these entities within the classification of lung cancer.

      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.

    • +

      GR03.04 - Molecularly-Defined Thoracic Malignancies (NUT, SMARCA4 and Others Sarcomas) (Now Available) (ID 3312)

      15:45 - 17:15  |  Presenting Author(s): Akihiko Yoshida

      • Abstract
      • Presentation
      • Slides

      Abstract

      Classification of tumors has been traditionally based on clinical and histological findings, with each entity often being characterized by molecular genetic changes. However, a few recently described tumor entities are defined by specific genetic abnormalities, and three such tumors are discussed here with a particular emphasis on their nosologic controversy. [NUT carcinoma] NUT carcinoma is a poorly-differentiated aggressive carcinoma with frequent squamous differentiation. NUT carcinoma harbors NUTM1 rearrangement by definition, with the most common fusion partner being BRD4 (~70%) and uncommon partners including NSD3 and BRD3. NUT carcinomas typically involve organs along the midline, such as the head, neck, and upper aerodigestive tract in young patients; however, a broader range of patient age and tumor sites exist. Recently, NUTM1 rearrangement has been reported in a small number of malignant tumors that lack epithelial differentiation, some of which show an undisputable phenotype of sarcoma. Tumors with CIC-NUTM1 fusion are the best known and their histological and transcriptomic similarities to CIC-DUX4 sarcomas suggest their relatedness with CIC sarcomas. Other NUTM1-rearranged sarcomas are highly heterogeneous, both histologically and genetically, including fusion partners such as BCORL1, MXD1, MXD4, and MGA. Interestingly, MGA-NUTM1 sarcomas have been repeatedly documented in the thoracic cavity of adults. More recently, NUTM1 rearrangement has been discovered in benign and malignant skin adnexal tumors. NUTM1 rearrangement is therefore no longer a signature of a single entity NUT carcinoma, and phenotypic correlation is critical for diagnosis. [SMARCA4-deficient thoracic sarcoma (DTS)] SMARCA4 is a core catalytic subunit of the SWI/SNF chromatin remodeling complex. SMARCA4 deficiency in thoracic tumors primarily occurs in association with carcinomas, accounting for 5–15% of lung adenocarcinomas and up to 30% of large cell and pleomorphic carcinomas. These carcinomas typically affect smoking men and are more common in poorly differentiated TTF1-negative tumors that are wild-type for EGFR and ALK. SMARCA4- DTS is a recently recognized sarcoma type with fewer than 60 cases reported to date. SMARCA4-DTS most commonly occurs in young to middle-aged adult men (median, 40 years old) with heavy smoking exposure and presents as large tumors in the thoracic cavity. SMARCA4-DTSs are aggressive, and the median survival is 4–7 months. Histologically, the tumors consist of diffusely infiltrating large dyscohesive epithelioid cells with relatively monotonous nuclei and prominent nucleoli, similar to proximal-type epithelioid sarcoma. Rhabdoid cells are seen in a subset of cases. By definition, all cases are deficient in SMARCA4 immunohistochemically because of inactivating SMARCA4 mutation. SMARCA4-DTS is different from SMARCA4-deficient lung carcinoma with respect to demographics (younger), clinical outcome (worse), histological features (more dyscohesive), immunophenotype (frequent positivity for CD34, SOX2, and/or SALL4, and negativity for claudin-4), and gene expression profiles. Interestingly, some SMARCA4-DTS tumors tested have frequent C:G/A:T transversion mutations and mutations in TP53, KRAS, KEAP1, and/or NF1, a shared profile with smoking-associated lung adenocarcinomas. The question has thus been raised whether these sarcomas might represent a dedifferentiated form of lung carcinoma. Nonetheless, an epithelial component has not been reported in any of the documented SMARCA4-DTS cases. Furthermore, most examples are not centered in the lung, and some entirely lack lung parenchymal involvement. [Primary pulmonary myxoid sarcoma (PPMS) with EWSR1-CREB1] PPMS is a rare low-grade lung sarcoma of young adults often presenting as an endobronchial mass. The tumor consists of multinodular myxoid growth that is populated by corded or reticular proliferation of spindle and/or epithelioid cells. These tumors often coexpress vimentin and epithelial membrane antigen and harbor EWSR1-CREB1 fusion. Tumors with a similar histological appearance have recently been reported in various soft tissue and visceral sites, including the brain, by the names of myxoid variant of angiomatoid fibrous histiocytoma (AFH) and intracranial myxoid mesenchymal tumors, which harbor EWSR1 fusions with genes encoding one of the CREB family transcription factors (ATF1, CREB1, or CREM). Primary pulmonary AFHs have been reported, with some showing myxoid features. Although PPMS is recognized in the WHO classification of the lung as a distinctive tumor, a significant overlap in histology and genetics, albeit several differences, may suggest a close relationship between PPMS and myxoid AFH.

      References:

      1. French CA. NUT Carcinoma: Clinicopathologic features, pathogenesis, and treatment. Pathol Int. 2018 Nov;68(11):583-595.

      2. Dickson BC, et al. NUTM1 Gene Fusions Characterize a Subset of Undifferentiated Soft Tissue and Visceral Tumors. Am J Surg Pathol. 2018 May;42(5):636-645.

      3. Le Loarer F, et al. Clinicopathologic Features of CIC-NUTM1 Sarcomas, a New Molecular Variant of the Family of CIC-Fused Sarcomas. Am J Surg Pathol. 2019 Feb;43(2):268-276.

      4. Stevens TM, et al. NUTM1-rearranged neoplasia: a multi-institution experienceyields novel fusion partners and expands the histologic spectrum. Mod Pathol.2019 Feb 5. [Epub]

      5. Sekine S, et al. Recurrent YAP1-MAML2 and YAP1-NUTM1 fusions in poroma and porocarcinoma. J Clin Invest. 2019 May 30;130. [Epub]

      6. Le Loarer F, et al. SMARCA4 inactivation defines a group of undifferentiated thoracic malignancies transcriptionally related to BAF-deficient sarcomas. Nat Genet. 2015 Oct;47(10):1200-5.

      7. Yoshida A, et al. Clinicopathological and molecular characterization of SMARCA4-deficient thoracic sarcomas with comparison to potentially related entities. Mod Pathol. 2017 Jun;30(6):797-809.

      8. Thway K, et al. Primary pulmonary myxoid sarcoma with EWSR1-CREB1 fusion: a new tumor entity. Am J Surg Pathol. 2011 Nov;35(11):1722-32.

      9. Smith SC, et al. At the intersection of primary pulmonary myxoid sarcoma and pulmonary angiomatoid fibrous histiocytoma: observations from three new cases. Histopathology. 2014 Jul;65(1):144-6.

      10. Schaefer IM, et al. Myxoid variant of so-called angiomatoid "malignant fibrous histiocytoma": clinicopathologic characterization in a series of 21 cases. Am J Surg Pathol. 2014 Jun;38(6):816-23.

      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.

    • +

      GR03.05 - Grading of NSCLCs - Problems and Solutions (Now Available) (ID 3313)

      15:45 - 17:15  |  Presenting Author(s): Andre Luis Moreira

      • Abstract
      • Presentation
      • Slides

      Abstract

      Background

      Tumor grading has been a traditional component of the pathologic evaluation and in many organ systems, tumor grading offers guideline to therapy and patient management. The latter has not been applied to NSCLC. However, considering the new advances in therapy modalities for NSCLC and advances in adenocarcinoma classification, it is now clear that there are different types of adenocarcinomas and these tumors should not be treated the same way. The 2015 WHO classification of pulmonary adenocarcinoma based on the predominant histological pattern has consistently been found to correlate with prognosis. There is broad agreement that the five histological patterns (lepidic, acinar, papillary, solid and micropapillary) are important prognostic indicators. Recent studies have proposed the inclusion of a number of additional pathologic features (the role of secondary patterns, non-traditional pattern such as cribriform and complex glandular patterns, nuclear grade, mitotic counts, presence of spread through alveolar space (STAS), and necrosis.) that also have prognostic value. The addition of these histological features to the predominant pattern could offer greater refinement of a grading scheme. Supplementing the classification of lung adenocarcinomas with an objective grading system will help define prognostic groups that could benefit from the changing landscape of emerging management and treatment options.

      Contrary to adenocarcinoma, there has been little advancements in the histological prognostic indicators in squamous cell carcinoma of the lung. Isolated reports have suggested that the presence of tumor budding into the stroma is the sole indicator of poor prognosis. Keratinization, which has been traditionally used to grade these tumors, does not appear to have prognostic value. However, a systematic evaluation of prognostic markers in these tumors have not been carried out. A summary of the current efforts in squamous cell carcinoma will be discussed.

      The IASLC pathology panel has proposed a systematic study to evaluate a set of histological criteria that have been described as prognostic indicators in adenocarcinoma aimed at establishing an objective grading system of invasive lung adenocarcinoma.

      Design

      A multi-institutional study involving well-annotated multiple cohorts of stage 1 adenocarcinomas with at least five years of follow up were evaluated. Annotation included an estimate of the percentage for each histological pattern present for each case; nuclear grade, cytology grade; and mitotic counts with pattern hot-spot association, presence of STAS, and necrosis. A cohort of 284 cases was used as a training set. Univariate analysis was performed to identify significant associations of histological features with recurrence-free survival and overall survival. ROC curve analysis was used to select the best model based on combinations of several features and its association with disease recurrence or death of disease. The results were validate on independent cohorts of 212 cases.

      Results

      Review of the literature showed that there are many variation in the classification and definitions on non-traditional patterns. In our cohorts, cribriform and complex glandular patterns followed similar curve as traditional high grade patterns (solid and micropapillary), therefore these non-traditional pattern were defined as patterns of high grade in the model. Another are of variation is the percentage of high grade pattern that can influence outcome. Therefore, the cut-off for a high grade pattern associated with recurrence or death of disease was also established in the training cohort and correspond to 20%. Therefore, amounts smaller than 20% of high grade pattern did not influence outcome.

      In the training cohort (n=284) all parameters tested, predominant patterns, mitotic count, nuclear grade, cytological grade, and STAS (but not necrosis) were found to have significant prognostic value on a univariate analysis. A Baseline Model composed of Age + Gender + Race + Type of surgery + Pathological Stage; showed an AUC of 0.673. In an attempt to improve this curve, histological parameters were added to the model.

      The addition of only the predominant pattern to the baseline increases the AUC to 0.698.

      A model based on the combination of predominant pattern paired with the second predominant pattern was found to have the highest AUC (0.765), followed by a combination of predominant pattern plus worse pattern (AUC=0.74). Addition of other histological features (nuclear grade, mitotic count, STAS etc.) did not significantly improve the model.

      Similar results were found in the validation set (N=212). The combination of the two most predominant patterns showed an AUC = 0.763, followed by a combination of predominant + worse pattern with AUC = 0.766. Addition of other histological features did not show improvement of the model.

      There was no statistical difference between the models using the two most predominant patterns and the predominant plus worse. There was good reproducibility scores for the 2 models

      Conclusion

      Our results suggest that an objective grading system for pulmonary adenocarcinoma is possible. Considering that there is no significant differences between a model that accounts for the 2 most predominant pattern and another composed of the predominant plus worse pattern. The IASLC pathology panel proposes the later to be used, because pathologists traditionally grade tumors by the worse component. Therefore, histologic assessment of the predominant pattern and worse pattern, would represent the most parsimonious and prognostic grading system for stage I lung adenocarcinomas.

      The use of the model in two other independent cohorts of adenocarcinomas (stages 1-3), as well as a reproducibility study will be discussed.

      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.



Author of

  • +

    MA10 - Emerging Technologies for Lung Cancer Detection (ID 129)

    • Event: WCLC 2019
    • Type: Mini Oral Session
    • Track: Screening and Early Detection
    • Presentations: 1
    • Now Available
    • +

      MA10.01 - Invasive Adenocarcinoma in Screen Detected Pure Ground-Glass Nodules (GGN) (Now Available) (ID 2736)

      15:15 - 16:45  |  Author(s): Ming Sound Tsao

      • Abstract
      • Presentation
      • Slides

      Background

      A major criticism of lung cancer screening initiatives is their propensity to instigate enhanced surveillance and over-treatment of otherwise indolent disease, including adenocarcinoma-in-situ (AIS). These nodules present radiographically as GGN. There are wide variations in the recommendations for surveillance (repeat imaging), diagnosis (biopsy) and therapeutic intervention (resection) for these lesions. To further our understanding of the optimal management of screen detected GGN, we used data from two screening studies in Canada with up to 17 years of follow-up to determine the proportion of persistent GGN that are invasive adenocarcinomas.

      Method

      Two lung cancer screening studies data sets were reviewed: the BC Lung Health Study (BCLHS) with 1365 participants and the Pan-Canadian Early Detection of Lung Cancer Study (PanCan) with 2537 participants. BCLHS enrolled ever smokers 45-74 years of age with >30-year smoking history. The median follow-up in this cohort was 12 years (0.1-17.6) The PanCan study screened participants age 50-75 years with a PLCOm2008 6-year lung cancer risk > 2%. The median follow-up was 5.5 years (3.2-6.1). The nodules were followed until they resolved, demonstrated stability for >2 yrs or were surgically resected. All pure GGO resected were re-reviewed and classified by two pulmonary pathologists according to the revised 2015 World Health Organization classification of lung tumours. Cancers were staged using the 8th edition of the AJCC/UICC cancer staging manual.

      Result

      A total of 18,589 nodules in 3902 participants were reviewed. 2392 (13% of all nodules) were classified as pure GGN. 1073 of the 2392 were > 5mm at the baseline scan. Of these 1073 GGN, 156 (15%) resolved, 879 (82%) remained pure GGN, 38 (3.5%) became part-solid or solid. 32(3%) of the GGN from 29 patients that demonstrated growth were resected. The median size prior to resection was 16 mm (range 7 to 33 mm). The histopathology distribution included: 19 invasive adenocarcinomas, 7 minimally invasive adenocarcinomas, 6 adenocarcinoma-in-situ. The TNM stage distribution and average size of the GGN on the CT prior to resection are listed in Table 1. Sixty-one percent of the invasive cancers (Stage IA1 to IIIA) were less than 20 mm. Eleven percent of the invasive adenocarcinomas had lymph node metastasis.

      presentation3.jpg

      Conclusion

      A high proportion of pure GGN that demonstrate growth are invasive cancers. The majority were < 20mm in size when they were resected. This has significant implication in the development of recommendations to manage screen detected GGN.

      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.

  • +

    MA11 - Immunotherapy in Special Populations and Predictive Markers (ID 135)

    • Event: WCLC 2019
    • Type: Mini Oral Session
    • Track: Immuno-oncology
    • Presentations: 1
    • Now Available
    • +

      MA11.04 - Platinum Doublet + Durvalumab +/- Tremelimumab in Patients with Advanced NSCLC: A CCTG Phase IB Study - IND.226 (Now Available) (ID 927)

      14:00 - 15:30  |  Author(s): Ming Sound Tsao

      • Abstract
      • Presentation
      • Slides

      Background

      Studies of single agent immune checkpoint inhibitors with platinum-based chemotherapy in non-small cell lung cancer (NSCLC) have demonstrated survival benefit over chemotherapy alone. The primary objective of this multi-centre study was to evaluate the safety and tolerability of durvalumab (Du), a PD-L1 inhibitor, +/- tremelimumab (Tr), a CTLA-4 inhibitor, with one of four standard platinum-doublet regimens (pemetrexed (pem), gemcitabine, etoposide (each with cisplatin or carboplatin) or nab-paclitaxel (with carboplatin)), in order to establish a recommended phase II dose (R2PD) for each combination. This abstract updates the results in the NSCLC cohort in this study.

      Method

      Patients (pts), regardless of tumour PD-L1 status, were enrolled into one of six dose levels (Table 1). Dose escalation was according to a Rolling Six type design. Concurrent enrollment of cohorts was allowed. ind 226 abstract wclc methods.png

      Result

      Seventy-three pts (median age=63 (range 34-80); 52% female; 77% non-squamous) were enrolled. The majority of drug-related adverse events (AEs) were grade 1 or 2. Most AEs were related to chemotherapy; other AEs were chemotherapy or immune-related (renal, hepatic, skin and pulmonary toxicity). AEs that were considered related to Du or Tr (immune related AEs (irAEs)) were mainly grade 1 or 2. The most common irAEs were fatigue (64%), rash/itch (42%), diarrhea/colitis (34%), anorexia (22%), thyroid dysfunction (19%), and nausea/vomiting (21/12%). The most common grade 3 or 4 irAEs were diarrhea/colitis (11%), fatigue (10%), and rash (5%). No treatment related grade 5 toxicities were reported. Twenty pts (27%) discontinued treatment due to an AE. Twelve pts (16%) discontinued treatment for toxicity related to D+/-T. Objective response rate (ORR) was 50.7% (95% CI = 38.7-62.6%). Median progression free survival (mPFS) was 6.5 months (95% CI = 5.5-9.4). Median overall survival (mOS) was 19.8 months (95% CI = 14.8-not yet reached). ORR was similar for all levels of PD-L1 staining including PD-L1 negative patients. ORR for pts with EGFR mutations (N=5) was similar to the ORR of wild type pts. Exploratory analyses suggest mPFS and mOS were longer in patients who experienced irAEs.

      Conclusion

      In this PD-L1 unselected patient population, Du and Tr can be safely combined with full doses of platinum-doublet chemotherapy. The ORR, mPFS and mOS are similar to results reported from other immunotherapy + chemotherapy combination trials. A randomized trial, CCTG BR.34, is evaluating the incremental benefit of adding platinum doublet to Du+Tr.

      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.

  • +

    MA18 - Advances in Diagnosis of Common Types of NSCLC (ID 145)

    • Event: WCLC 2019
    • Type: Mini Oral Session
    • Track: Pathology
    • Presentations: 1
    • Now Available
    • +

      MA18.07 - Identification of Neuroendocrine Transformation in Anaplastic Lymphoma Kinase Rearranged (ALK+) Tumors After Tyrosine Kinase Inhibitors (Now Available) (ID 1137)

      11:30 - 13:00  |  Author(s): Ming Sound Tsao

      • Abstract
      • Presentation
      • Slides

      Background

      Acquired resistance after ALK tyrosine kinase inhibitors treatment has multiple known mechanisms: new mutations or gene amplifications, bypass signaling and rarely neuroendocrine histological transformation. Here we describe results of a program utilizing routine biopsy post-progression in ALK+ patients for clinical and research purposes.

      Method

      Since 2014, ALK+ lung cancer patients treated at the Princess Margaret Cancer Centre have undergone routine biopsies at disease progression time points upon failure of an ALK-tyrosine kinase inhibitor (TKI) for both clinical purposes and research purposes, in particular to obtain tissue for primary derived xenograft (PDX) engraftment.

      Result

      All 9/9 patients consented for research sampling during clinical biopsy procedures (median 2 extra cores/passes); 2 patients were biopsied more than once; 3 PDX models from 2 patients have engrafted; 3 additional models are too early to assess engraftment. Engraftment occurred in patients with clinically aggressive tumors and poor survival outcomes. In this process, we identified 2 patients with neuroendocrine transformation post-second generation ALK TKI: (a) a 59 yo Asian female, never smoker, diagnosed six years prior with metastatic disease, heavily pretreated with crizotinib (12 months), pemetrexed (16 months), ceritinib (25 months), alectinib (6 months) and brigatinib (3 months); post-alectinib biopsy showed no transformation, while post-brigatinib liver biopsy demonstrated transformation to large cell neuroendocrine carcinoma; (b) a 75 yo Caucasian female, never smoker, diagnosed eight months prior and started on alectinib with a partial response, progressed in a single site; endobronchial biopsy demonstrated high grade neuroendocrine transformation. Both biopsies were positive for neuroendocrine markers (chromogranin and synaptophysin), TTF-1 and diffusely co-expressed ALK on immunohistochemistry. Assessment of PDX engraftment of these models is ongoing.

      Conclusion

      Routine combined clinical and research biopsy of ALK+ patients at time of TKI failure helped to identify these recent cases of neuroendocrine transformation as a possible mode of resistance and provide tissue for model development. This is the first time that ALK+ transformation to large cell neuroendocrine carcinoma is reported in the literature. (PP, AFF, SNMF, LN contributed equally).

      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.

  • +

    OA08 - Advanced Models and "Omics" for Therapeutic Development (ID 133)

    • Event: WCLC 2019
    • Type: Oral Session
    • Track: Biology
    • Presentations: 1
    • Now Available
    • +

      OA08.01 - Organoid Cultures as Novel Preclinical Models of Non-Small Cell Lung Cancer (Now Available) (ID 2115)

      11:00 - 12:30  |  Presenting Author(s): Ming Sound Tsao

      • Abstract
      • Presentation
      • Slides

      Background

      There is an unmet need to develop novel clinically relevant models of NSCLC to accelerate identification of drug targets and our understanding of the disease. Organoids, which are cells grown in three-dimensional environments in Matrigel, have emerged as novel preclinical models of cancer. Recently protocols for generating NSCLC organoids have been reported, but the growth, and molecular features of organoids as compared to their matching primary patient tumor or patient-derived xenografts (PDX) remain vague.

      Method

      Thirty surgically resected NSCLC patient tumor and 35 PDX tissue of lung adenocarcinoma and squamous cell carcinoma subtypes were processed for organoid establishment. Organoids and matching tumor tissues were characterized by histology and immunohistochemistry, and molecularly profiled by whole exome and RNA-sequencing. Subcutaneous injection of organoids in vivo was performed to confirm tumorgenicity. Organoids were subjected to drug testing and drug response was verified in the matched PDX.

      Result

      Using a novel culture condition that our laboratory developed, we have collected tumor samples from 16 primary and 13 PDX samples of adenocarcinoma (n=29) and 14 primary and 22 PDX samples of squamous cell carcinoma (n=36). Over 85% (57/65) of our patient and PDX tumor tissues formed organoids that exhibited a wide range of short-term (<3 months) and long-term (>3 months) growth. Specifically, the success rate of establishing short-term and long-term models are 74% (48/65) and 14% (9/65), respectively. The long-term propagable organoids recapitulated the histology of the patient and PDX tumor. They also retained the ability to form xenograft in NOD-SCID mice. The organoids preserved mutation, copy number aberrations and global gene expression profile of the parental tumors. We additionally showed the utility of short-term and long-term organoids for identifying biomarkers of sensitivity to drugs and combinational targeted therapies.

      Conclusion

      NSCLC organoids are novel patient-derived ex-vivo tumor models for anti-cancer drug screening and biomarker discovery, thus could be incorporated into novel drug discovery pipelines. Further efforts are ongoing to increase the success rate of establishing long-term organoid lines.

      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.

  • +

    OA13 - Ideal Approach to Lung Resection and Novel Perioperative Therapy (ID 146)

    • Event: WCLC 2019
    • Type: Oral Session
    • Track: Treatment of Early Stage/Localized Disease
    • Presentations: 1
    • Now Available
    • +

      OA13.01 - SPECS2 Lung Cancer Consortium Prospective Multicenter Validation of Prognostic Signature for Early Stage Squamous Lung Cancer (Now Available) (ID 2723)

      11:30 - 13:00  |  Author(s): Ming Sound Tsao

      • Abstract
      • Presentation
      • Slides

      Background

      Squamous Lung Cancer (SC) which constitutes 30% of all non-small cell lung cancers (NSCLC) has few targeted therapy options for advanced disease. Surgery for early SC is the best treatment strategy; however, even patients who undergo surgery for stage IA or IB disease are still at a substantial risk for recurrence and death. Adjuvant therapy is not currently indicated for stage I SC smaller than 4 cm. Prior reports suggest gene expression-based signatures that may predict recurrence in patients with stage I SC, but none has been validated or is in clinical use. The SPECS2 Lung Cancer Consortium was assembled to compare and attempt to validate previously published prognostic signature(s) according to the guidelines proposed by Subramanian and Simon (J Natl Cancer Inst 2010; 7:327).

      Method

      The multi-institutional team assembled 249 frozen SC samples representing six participating institutions (cohort 1). These samples were fully annotated in a redcap database hosted by the independent statistical core. Cohort 2 was assembled utilizing 234 frozen SC samples from a prospective multi-institutional NCTN lung biobanking protocol (NCT00899782). RNA was extracted and profiled with U133A microarrays (Affymetrix) in independent core facilities. The data was transferred directly to the SPECS2 Lung statistical core in collaboration with the Alliance Statistical core and the performance of 6 most promising candidate signatures was evaluated relative to a base model that included only age, gender and AJCC stage (editions 6, 7, 8).

      Result

      Analysis of Cohort 1 demonstrated that only one signature (Raponi et al, Cancer Res 2006; 66:7466) significantly enhanced prognosis relative to the base model, independent of AJCC edition. This was also observed in Cohort 2, where Uno’s C index associated with AJCC 8th edition stage, sex and age (0.561; 0.468-0.654) was significantly (p <0.05) increased when the prognostic signature was added to the model (0.683; 0.611-0.755).

      Conclusion

      The SPECS2 Lung Cancer Consortium was successful in validating a previously published prognostic molecular signature for early stage SC using rigorous experimental design. To our knowledge, this is the first unbiased validation of a lung cancer prognostic signature using multi-institutional prospective specimens. These results support a clinical trial designed to evaluate the potential role of adjuvant therapy in completely resected early stage SC.

      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.

  • +

    P1.01 - Advanced NSCLC (ID 158)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Advanced NSCLC
    • Presentations: 2
    • Now Available
    • Moderators:
    • Coordinates: 9/08/2019, 09:45 - 18:00, Exhibit Hall
    • +

      P1.01-30 - Non-Small Cell Lung Cancer (NSCLC) Next Generation Sequencing (NGS): Integrating Genomic Sequencing into a Publicly Funded Health Care Model (Now Available) (ID 2588)

      09:45 - 18:00  |  Author(s): Ming Sound Tsao

      • Abstract
      • Slides

      Background

      Standard of care (SOC) molecular diagnostics for stage IV NSCLC patients in Ontario, Canada includes publicly reimbursed EGFR/ALK, and BRAF/ ROS-1 testing in selected cases. Other genomic alterations are not tested routinely at all institutions; however, enhanced molecular testing may broaden treatment options for patients by identifying actionable targets. This study evaluated costs, identified actionable targets, and determined clinical trial eligibility as a result of using the Oncomine Comprehensive Assay v3 (OCA v3, ThermoFisher) NGS in stage IV NSCLC patients at a single institution.

      Method

      This prospective study of stage IV NSCLC out-patients at Princess Margaret Cancer Centre (Toronto) began in February 2018 and recruitment is ongoing (NCT03558165). NSCLC patients without EGFR/ALK/KRAS/BRAF alteration (unless failure of prior targeted therapy and tissue rebiopsy), had diagnostic samples tested by OCAv3 (ThermoFisher; 161 genes: hotspots, fusions, and copy number variations). Primary endpoints were identification of incremental actionable targets and clinical trial opportunities as a result of broader OCAv3 testing. Secondary endpoints include feasibility and cost from the Canadian public healthcare perspective.

      Result

      From Feb 2018- Jan 2019 65 patients were enrolled [62% (N=40) completed/ 21% (N=14) screen fail/ 17% (N=11) pending], median age of completed cohort was 65, 60% (N=24) female, never/light smokers 68% (N=27), Asian 38% (N=15), previously treated 33% (N=13). Actionable targets beyond SOC were identified in 33% (N=13): ERBB2 (N=8), BRAFV600 (N=3), NRG fusion (N=1), MET exon 14 (N=1). Failure of NGS was secondary to insufficient tissue. 91% (N=10) of screen failures was secondary to tissue exhaustion from prior sequential SOC molecular testing. New clinical trial options were identified in 70% as a result of OCA v3 testing. Incremental costs per case beyond EGFR/ALK are estimated at $540 CAD. If ROS-1 and BRAF testing were publicly reimbursed at current rates, the incremental profiling cost with OCAv3 would be $90 CAD per case.

      Conclusion

      The OCAv3 consolidates genomic testing, identifies additional actionable targets, and substantially increases clinical trial eligibility for patients at a small incremental cost. Sample failures are reflective of exhausted diagnostic tissue as a result of prior sequential genomic testing. The key barrier to implementation of NGS remains funding in the Canadian health care system.

      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.

    • +

      P1.01-70 - Dominant Circulating Myeloid Populations Are Associated with Poor Response in NSCLC Treated with 1st Line PD-1 Monotherapy (Now Available) (ID 2295)

      09:45 - 18:00  |  Author(s): Ming Sound Tsao

      • Abstract
      • Slides

      Background

      Immune subpopulations within the tumor microenvironment (TME) play a central role in determining response to checkpoint inhibitors. Myeloid derived suppressor cells (MDSC), a heterogeneous population of immature myeloid cells, have a predominantly immunosuppressive role by stimulating T regulatory cells. We hypothesize that elevated myeloid-to-lymphocyte measures in the peripheral blood predict for greater numbers of myeloid derived suppressor cells in the TME and worse outcomes.

      Method

      We identified all advanced NSCLC patients treated with immunotherapy between 2010-2019 at the Princess Margaret Cancer Center. Patients who received first line monotherapy with a PD-1 inhibitor were reviewed for clinical information including age, sex, histology, stage, smoking status, ethnicity, PD-L1 expression and tumor genotype. Myeloid cells lines analyzed included neutrophils, monocytes and platelets, expressed as ratios to peripheral lymphocytes. Multivariate analyses were conducted using the cox and logistic regression models to adjust for confounders.

      Result

      We identified 75 patients who were eligible for analysis. Disproportionate increases in the different myeloid cell types were highly correlated with each other (all Pearson’s rho>0.8) and the neutrophil to lymphocyte ratio (NLR) was selected as representative. A high NLR (>5) was associated with shorter time-to-treatment-failure (median TTF 9.7 vs 29.4 months) that remained significant after adjusting for confounders including PD-L1 and presence of liver metastases (p=0.004). High NLR was also an independent predictor of poor OS (median 11.3 vs 56.8 months, HR 3.02, p=0.04). Although NLR was not predictive of radiographic response, there was a trend to association with a rapidly progressive phenotype defined by primary progressive disease and a duration of therapy ≤2 months (p=0.06). Other predictive factors included the presence of liver metastases, which was associated with a worse OS (HR3.37 p=0.05) but not TTF (p=0.14). An association was also seen between NLR and liver metastases (mean NLR 6.6 vs 25.2 in the absence and presence of liver metastases respectively, p<0.001).

      Conclusion

      A disproportionate increase in peripheral immune myeloid populations may represent a systemic, myeloid-driven, immunosuppressive state that is significantly associated with primary refractory disease, rapid progression, and poor survival. A subset of about 50 patients with biobanked tissue are presently being analyzed using multiplex immunofluorescence to assess for MDSCs in the TME to correlate with peripheral blood findings.

      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.

  • +

    P1.14 - Targeted Therapy (ID 182)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Targeted Therapy
    • Presentations: 1
    • Moderators:
    • Coordinates: 9/08/2019, 09:45 - 18:00, Exhibit Hall
    • +

      P1.14-07 - Genomic Profiling of Liquid Biopsies During 2nd/3rd Generation ALK Inhibitor Therapy to Identify Novel Mechanisms of Resistance (ID 804)

      09:45 - 18:00  |  Author(s): Ming Sound Tsao

      • Abstract

      Background

      Second- and third-generation ALK inhibitors each have diverse mechanisms of resistance. Only a fraction of resistance is due to secondary mutations of the ALK gene. Altered bypass tracts are likely the case in some other instances. Genomic alterations of other genes and pathways may be a third mechanism of resistance. Repeat liquid biopsies during the course of patients’ treatments can provide a minimally invasive method for sampling cancer-specific genomic information that leads to improved treatment selection.

      Method

      In the Lung Cancer Clinic of the Princess Margaret Cancer Centre, serial plasma samples were collected from six lung cancer patients with ALK rearrangement at multiple serial clinic visits pre- and post- progression on next-generation ALK inhibitors. We focused on next generation agents, as there has been previous focus on crizotinib resistance mechanisms already. Cell-free DNA (cfDNA) was extracted (median: 50 ng; range: 20-2760 ng) and profiled using a next-generation sequencing (NGS) platform with Geneseeq Prime 425-gene panel at a mean coverage depth of 4747X (and a deduplicated mean coverage depth of 2160X).

      Result

      Somatic alterations from plasma cfDNA were detected in all six patients at various time points with three patients having detectable ALK alterations. Systemic progression (2/2 patients) correlated well with the ability of liquid biopsies to detect somatic mutations, while central nervous system (CNS)-predominant progression did not (4/4 patients). One patient, after disease progression on ceritinib, alectinib and brigatinib, exhibited variable allele fractions (AFs) of ALK G1202R mutation in cfDNA. The levels of G1202R decreased and ultimately became undetectable, corresponding to the patient’s clinical response to lorlatinib. In a patient who exhibited significant systemic progression, a massive increase in mutation AFs and many newly acquired mutations were detected in the cfDNA, including NOTCH1, DICER1, BRCA2, TP53, CDKN2A, ERBB3, and FAT1mutations. However, the increase in the number of co-mutations was not related to increases in the amount of extracted cfDNA.

      Conclusion

      Broad panel-based NGS of plasma cfDNA enabled noninvasive detection of systemic (but not CNS-predominant) progression during second and subsequent generation ALK inhibitor treatment, and can identify known and putative mechanisms of resistance for treatment decision-making.

  • +

    P2.03 - Biology (ID 162)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Biology
    • Presentations: 2
    • Moderators:
    • Coordinates: 9/09/2019, 10:15 - 18:15, Exhibit Hall
    • +

      P2.03-11 - Impact of Ethnicity on Outcome in Never Smokers with EGFR and ALK Wildtype (EGFR/ALK-Wildtype) Lung Adenocarcinomas (ID 2035)

      10:15 - 18:15  |  Author(s): Ming Sound Tsao

      • Abstract

      Background

      EGFR-mutations and ALK-rearrangements are frequent in lung adenocarcinoma (LUAD) samples from never smoker patients. Nevertheless, up to a quarter of all LUAD cases in never smokers are EGFR/ALK-wildtype: these patients have limited therapeutic options and few well-established clinical and molecular predictors of outcome. Our main objectives here were to investigate the prognostic impact of ethnicity in never smoker patients with EGFR/ALK-wildtype LUAD and seek for specific somatic events correlated to ethnical background in these patients.

      Method

      We included 85 samples from lifetime never-smoker patients with EGFR/ALK-wildtype LUAD collected from surgical resection with curative intent. Stages 1/2/3 were identified in 56 (66%)/15 (18%)/14 (16%) samples. A subset of those samples (n=46), with similar stage distribution, had snap-frozen tumor and paired-adjacent tissue available and were submitted to paired-end whole-exome sequencing. Fisher’s exact and Chi-squared tests were used to compare specific mutations between Asians vs non-Asians. Recurrence-free-survival (RFS) was calculated based on the Kaplan-Meier method; Cox modeling was used to generate hazard ratios (HR), adjusted for key clinical features.

      Result

      Most patients in the cohort were female (63/85, 74%); the median age was 68 years; median follow-up was 51 months. According to self-reports, 19/85 (22%) and 66/85 (78%) patients identified as Asians and non-Asians, respectively; no major clinical and pathologic differences were identified between these populations. Five-year recurrence free survival was significantly lower for Asians compared to non-Asians (50% vs. 78%, adjusted HR = 2.9; CI = 1.1-7.8, p=0.02), Figure 1. Among somatic events, in-frame deletions in CNPY3 (Toll-like receptor-specific co-chaperone for HSP90B1) were more frequent in Asians (30%) compared to non-Asians (18%). In contrast, DDX11 missense mutations (21% vs 0%; nucleic acid binding protein involved in genome stability), NOTCH2 multi-hits and frame-shift deletions (7% vs 1%), and KRAS missense mutations (7% vs 0%) were more frequently altered in non-Asians than in Asians.

      Conclusion

      In our cohort of never-smoker patients with EGFR/ALK-wildtype LUAD, Asian patients showed higher relapse rates than non-Asians. We identified differentially mutated genes by ethnicity that may partly account for these differences in outcome. (SNMF and AFF contributed equally)

      figure 1.jpg

    • +

      P2.03-37 - Genomic Landscape of EGFR/ALK Wild-Type Lung Adenocarcinomas in Never-Smokers and Importance of Epithelial-Mesenchymal-Transition (ID 1283)

      10:15 - 18:15  |  Author(s): Ming Sound Tsao

      • Abstract

      Background

      The molecular landscape of EGFR/ALK wild-type Lung Adenocarcinomas in never-smokers is poorly understood. Never-smokers usually have low PD-L1 expression and low Tumor Mutation Burden, challenging treatment strategies when no known driver-mutations are found. To identify putative driver mutations, we compared whole exome sequencing (WES) results in the EGFR/ALK wild-type Lung Adenocarcinoma in the never smokers Toronto cohort with a corresponding EGFR/ALK wild-type Lung Adenocarcinoma group of smokers from TCGA.

      Method

      For never-smokers with resected EGFR/ALK wild-type Lung Adenocarcinomas, frozen tumor and paired-normal-lung were evaluated by WES at a mean coverage of 238x. The paired-end reads were aligned using BWA and were further processed using the standard GATK pipeline. Somatic mutations and indels were identified using MuTect and VarScan, respectively. We compared mutations from our cohort to the TCGA smokers who had EGFR/ALK wild-type Lung Adenocarcinomas from publicly available data (TCGA) to identify genes at least 10% more frequently mutated in never smokers compared to the TCGA cohort.

      Result

      In our cohort with 45 never-smoker patients, 80% were females; median age was 70y; 29% were Asians; Stage I/II/III+ were 71%/15%/13%; after a median follow-up of 69 months, 24% had recurred. Median non-synonymous Tumor Mutation Burden was 1.3mut/Mb in never-smokers. We identified 39 genes that were more frequently mutated in never-smokers vs smokers, including some known tumor suppressor genes. The most prevalent genes included ADAM21 missense mutations (21% vs 1%; adj p=0.003), NOTCH2 frame-shift deletions and multi-hit mutations (40% vs 17%; adj p=0.04), MST1 missense mutations and in-frame deletions (13% vs 0%; adj p=0.008), ZMIZ2 frame-shift insertions (13% vs 0%; adj p=0.008) and FOXD4 missense mutations (10% vs 0%; adj p=0.02). Many of these differentially mutated genes have been previously associated to epithelial-mesenchymal-transition signaling pathways. Conversely and as expected, KRAS, TP53, STK11 and KEAP1 were more frequently mutated in the TCGA smokers EGFR/ALK wild-type Lung Adenocarcinomas cohort.

      oncoprint wes.png

      Conclusion

      We identified multiple genes, particularly involved in the epithelial-mesenchymal-transition signaling pathways that are over-represented in never-smokers with EGFR/ALK wild-type Lung Adenocarcinomas, when compared to smokers with EGFR/ALK wild-type Lung Adenocarcinomas. This is a novel finding with potential clinical importance. Validation studies, analyzing epithelial-mesenchymal-transition signaling activation pathways on the EGFR/ALK wild-type Lung Adenocarcinomas never smokers population are needed to best identify the actual role in carcinogenesis and metastasis, guiding future treatment strategies. (AFF and SNMF contributed equally).

  • +

    P2.09 - Pathology (ID 174)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Pathology
    • Presentations: 1
    • Moderators:
    • Coordinates: 9/09/2019, 10:15 - 18:15, Exhibit Hall
    • +

      P2.09-24 - IASLC Global Survey for Pathologists on PD-L1 Testing for Non-Small Cell Lung Cancer (ID 906)

      10:15 - 18:15  |  Author(s): Ming Sound Tsao

      • Abstract
      • Slides

      Background

      PD-L1 immunohistochemistry (IHC) is now performed for advanced non-small cell lung cancer (NSCLC) patients to examine their eligibility for pembrolizumab treatment, as well as in Europe for durvalumab therapy after chemoradiation for stage III NSCLC patients. Four PD-L1 clinical trial validated assays (commercial assays) have been FDA/EMA approved or are in vitro diagnostic tests in multiple countries, but high running costs have limited their use; thus, many laboratories utilize laboratory-developed tests (LDTs). Overall, the PD-L1 testing seems to be diversely implemented across different countries as well as across different laboratories.

      Method

      The Immune biomarker working group of the IASLC international pathology panel conducted an international online survey for pathologists on PD-L1 IHC testing for NSCLC patients from 2/1/2019 to 5/31/2019. The goal of the survey was to assess the current prevalence and practice of the PD-L1 testing and to identify issues to improve the practice globally. The survey included more than 20 questions on pre-analytical, analytical and post-analytical aspects of the PDL1 IHC testing, including the availability/type of PD-L1 IHC assay(s) as well as the attendance at a training course(s) and participation in a quality assurance program(s).

      Result

      344 pathologists from 310 institutions in 64 countries participated in the survey. Of those, 38% were from Europe (France 13%), 23% from North America (US 17%) and 17% from Asia. 53% practice thoracic pathology and 36%, cytopathology. 11 pathologists from 10 countries do not perform PD-L1 IHC and 7.6% send out to outside facility. Cell blocks are used by 75% of the participants and cytology smear by 9.9% along with biopsies and surgical specimens. Pre-analytical conditions are not recorded in 45% of the institutions. Clone 22C3 is the most frequently used (61.5%) (59% with the commercial assay; 41% with LDT) followed by clone SP263 (45%) (71% with the commercial assay; 29% with LDT). Overall, one or several LDTs are used by 57% of the participants. A half of the participants reported turnaround time as 2 days or less, while 13% reported it as 5 days or more. Importantly, 20% of the participants reported no quality assessment, 15%, no formal training session for PD-L1interpretation and 14%, no standardized reporting system.

      Conclusion

      There is marked heterogeneity in PD-L1 testing practice across individual laboratories. In addition, the significant minority reported a lack of quality assurance, formal training and/or standardized reporting system that need to be established to improve the PD-L1 testing practice globally.

      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.

  • +

    P2.14 - Targeted Therapy (ID 183)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Targeted Therapy
    • Presentations: 2
    • Moderators:
    • Coordinates: 9/09/2019, 10:15 - 18:15, Exhibit Hall
    • +

      P2.14-40 - Tumor-Stromal Microenvironment Interactions in a PDX Model of EGFR TKI Drug Tolerance (ID 509)

      10:15 - 18:15  |  Author(s): Ming Sound Tsao

      • Abstract

      Background

      Sixty to eighty percent of advanced stage lung adenocarcinoma patients with sensitizing epidermal growth factor receptor (EGFR) mutated tumors respond to EGFR tyrosine kinase inhibitors (TKIs). However, the vast majority of patients eventually progress due to acquired resistance. In vitro evidence suggests that minor populations of drug tolerant cells (DTCs) may be important for tumors surviving TKI. These studies cannot investigate changes in non-cancerous cell populations found within tumors. Yet, stromal cells have been implicated in protecting cancer cells from treatment-induced death and early stage lung adenocarcinomas responding to neoadjuvant EGFR TKI exhibited DTCs within large areas of fibrosis (NCT00188617). We hypothesize that molecularly characterizing DTCs in vivo in comparison to an untreated tumor in a patient-derived xenograft (PDX) model may delineate stromal changes that sustain DTCs, and potentially mimic clinical events.

      Method

      DTCs were harvested after one month of chronic erlotinib exposure in a lung adenocarcinoma PDX model harboring an exon 19 deletion; an untreated baseline (BL) tumor was also harvested. Histological characterization and single-cell RNA-sequencing (scRNA-seq) of DTCs and BL tumors were compared. ScRNAseq cell-types were assigned using reference component analysis. RNA expression levels of receptor/ligands were explored in cell populations.

      Result

      Post-erlotinib treatment, cell-type proportions within the tumor shifted dramatically, with substantially fewer cancer cells and more fibroblasts, mesenchymal stem cells (MSCs), and natural killer cells (NKCs). Two antigen presenting cell transcriptomic states (APC1 and APC2) were identified in both DTC and BL tumors: APC1s exhibited translation-related gene expression profiles while APC2s exhibited immune-response profiles. BL tumors contained mostly APC1s, whereas DTC tumors exhibited more equal proportions of both APC types. Expression profiles for some cell-types also shifted after treatment. Fibroblasts and NKCs exhibited shifts toward more inflammatory and immune-responsive expression profiles post-treatment. Fibroblasts and endothelial cells demonstrated gene expression shifts towards decreased angiogenesis and vasculature development. Paired ligand-receptor interactions between cancer-stromal cells were increased or decreased congruently post-treatment. Specifically, fibroblasts exhibited a shift from alpha-SMA+ myofibroblastic to more IL6+ inflammatory phenotypes, by mRNA and equivalent immunohistochemistry, post-treatment. Cancer cells exhibited a reciprocal increase in IL6R receptor expression post-treatment.

      Conclusion

      Using an EGFR mutant PDX model sensitive to EGFR TKIs, we see substantial post-treatment changes after chronic TKI exposure in non-cancerous (stromal) cell population composition involving their proportions, expression profiles, and their inferred communication with cancer cells. Understanding these potentially protective shifts in non-cancerous cell populations post-TKI-treatment may help identify clinically-relevant mechanisms of drug tolerance.

    • +

      P2.14-62 - Early, Subclinical SCLC Transformation in Patients with EGFR Mutant Lung Cancer Receiving Osimertinib, Detected Through Cell-Free DNA (ID 812)

      10:15 - 18:15  |  Author(s): Ming Sound Tsao

      • Abstract

      Background

      Liquid biopsies provide a convenient approachfor serial sampling and real-time disease monitoring, leading to the early detection of treatment response, disease progression and drug-resistance. Here,we present genomic profiling of serial liquid biopsies from seven lung cancer patients with activatingEGFRmutations receiving osimertinib in clinical practice.

      Method

      At Princess Margaret Cancer Centre, in the Lung Cancer Outpatient Clinics, plasma samples were obtained from each patient at defined clinical visits (between ~1–5 months in-between visits). Cell-free DNA (cfDNA; with a median of 57 ng; range: 3.5 to 3806 ng) was extracted from plasma samples and profiled using targeted capture next-generation sequencing with the Geneseeq Prime 425-gene panel, at a mean coverage depth of 4892X (with a deduplicated mean coverage depth of 2108X).

      Result

      Systemic tumour burden correlated with the detection of genomic alterations in cfDNA: Four of four of the patients with low tumour burden, despite minor disease progression, exhibited minimal EGFR and co-mutation allele frequencies (AFs). Conversely, significant increases in systemic (but not central nervous system) tumour burden led to increases in driver and co-mutation AFs (two our of three patients). EGFR C797S mutation and inactivating mutations in RB1 and TP53 were detected in the cfDNA of one patient nearly four months prior to the development of small cell lung cancer (SCLC) transformation confirmed on tissue biopsy with distinct transformed and untransformed areas. Both of the specific RB1 and TP53 mutations found in cfDNA have been previously associated with SCLC. Subsequent combination cisplatin-etoposide chemoradiation resulted in temporary complete remission of the transformed SCLC, corresponding to loss of RB1 mutation detection by cfDNA testing.

      Conclusion

      Profiling of plasma cfDNA using hybrid capture deep sequencing of a large gene panel can detect early subclinical transformation of EGFR-mutated lung cancer into small cell lung cancer (i.e., neuroendocrine transformation), leading to earlier diagnosis and management of the transformed disease. Serial liquid biopsy profiling can also be used to monitor disease progression. However, detection sensitivity of tumour cfDNA largely depends on systemic tumour burden.

  • +

    P2.16 - Treatment in the Real World - Support, Survivorship, Systems Research (ID 187)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Treatment in the Real World - Support, Survivorship, Systems Research
    • Presentations: 1
    • Moderators:
    • Coordinates: 9/09/2019, 10:15 - 18:15, Exhibit Hall
    • +

      P2.16-05 - Population-Based ROS1 Testing in Lung Cancer: Creating Opportunity in a Publicly Funded System (ID 1480)

      10:15 - 18:15  |  Author(s): Ming Sound Tsao

      • Abstract
      • Slides

      Background

      ROS1 gene rearrangement is found in 1-2% of all non-small cell lung cancer (NSCLC) and is recommended as standard molecular diagnostic testing. This study models the most efficient ROS1 testing strategy to maximize detection of true positive cases (TP) while minimizing costs and turnaround time (TAT).

      Method

      A population-based ROS1 testing model was developed from a Canadian (Ontario) public healthcare system perspective examining the use of immunohistochemistry (IHC) and next generation sequencing (NGS) versus fluorescence in situ hybridization (FISH, gold standard), reflex versus molecular or clinical selection (never smokers), and blood- versus tissue-based testing. Model inputs were derived from existing literature and expert opinion. Direct testing costs and TAT were obtained from the Ontario public perspective (University Health Network, Cancer Care Ontario).

      Result

      The most cost-effective strategy for the outcomes of TAT and TP was reflex testing with IHC and subsequent FISH confirmation; this identified a high proportion of TP within a relatively short TAT. The most costly reflex strategy was NGS, with a greater proportion of missed TP (Table), and long TAT. Clinician selection of never smokers yielded the lowest proportion of TP. Population-based plasma ctDNA testing using commercial assays was the most costly strategy. One-way sensitivity analysis demonstrated that the TP outcome was most sensitive to the population prevalence of ROS1, while cost was most sensitive to the specificity of ROS1 IHC.

      Conclusion

      Pathologist-initiated reflex testing using IHC with FISH confirmation provides the most cost-effective population-based testing strategy. Clinician-initiated testing significantly lengthens TAT. Selecting only never smokers for testing misses a larger proportion of TP patients who would benefit from targeted therapy despite potential cost savings.

      table.png

      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.

  • +

    YI03 - Scientific Mentoring (ID 109)

    • Event: WCLC 2019
    • Type: Young Investigator Session
    • Track: Young Investigators
    • Presentations: 1
    • Now Available
    • +

      YI03.02 - How to Apply for International Fellowship? (Now Available) (ID 3704)

      10:30 - 12:00  |  Presenting Author(s): Ming Sound Tsao

      • Abstract
      • Presentation
      • Slides

      Abstract

      International fellowship to pursue further training is an important part of career development for junior clinicians and investigators. International fellowship will not only increase the clinical and/or research skill of the fellow, it also offers an important opportunity to establish interaction network with senior investigators and peer young investigators across the world. When applying for international fellowship, several aspects should be considered: (1) choosing the field for further training, (2) application procedure, and (3) funding availability. The choice of field for further training will largely determine the future career of the candidate and should be in line with one’s scientific or career passion and clinical/research interest. Prior experience with a role model and mentorship during earlier formative years usually has great influence on one’s decision to pursue career choice. Fellowship training can be limited to clinical only, research only or combined clinical and research training; the latter for a candidate who wishes to pursue clinician-scientist career. Once a decision to pursue further training is made, the candidate should look for fellowship opportunities. This is often through discussion with local mentors or colleagues with prior international fellowship experience, or via familiarity with potential mentors who have published extensively in the area of candidate’s interest or have lectured at international meetings. On-line searches for institutions with established fellowship program may also be useful, but many training programs may not be openly advertised, thus personal approaches and recommendations are often more fruitful. Writing an application letter that demonstrates strong background, qualification, prior track record, commitment and clear post-fellowship career pathway are key elements to win an opportunity for further interview. Availability of secured partial/full funding from local Institution or independent funding agency is a great asset for an application. While one year fellowship that involves only clinical training may be sufficient, training that involves laboratory research will usually require a minimum of 2 years tenure, as the latter usually requires more time to accomplish specific research projects. In such case, preliminary data obtained during first fellowship year may greatly contribute to applications for fellowship offered by international societies (e.g., IASLC, ASCO, AACR) or funding agencies.

      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.