Virtual Library

Start Your Search

A.L. Moreira

Moderator of

  • +

    MS21 - Practical Problems in Lung Cancer Diagnosis - Application of the 2011 Adenocarcinoma Classification (ID 38)

    • Event: WCLC 2013
    • Type: Mini Symposia
    • Track: Pathology
    • Presentations: 5
    • +

      MS21.1 - Cytological Diagnosis (ID 556)

      14:00 - 15:30  |  Author(s): K. Geisinger

      • Abstract
      • Presentation
      • Slides

      Abstract
      Whenever a major alteration in histopathologic classification occurs, important ramifications for cytopathology follow. Furthermore, the recent recognition that associations of specific types of epithelial malignancies of the lung are associated with different prognoses and therapeutic complications also directly impacts diagnostic cytology. One important question is: how well can we distinguish adenocarcinoma, squamous cell carcinoma, and other nonsmall cell carcinomas (NSCLC) from each other in cytologic preparations? The answer is very well, with most recent data emanating from aspiration samples. Based purely on routine cytomorphology, approximately 90% of all specimens with NSCLC can be rendered by strict adherence to classic well indoctrinated cellular features. When the distinction cannot be rendered by morphology alone, the addtion of a small battery of immunochemical reactions raises the proportion of correct cell typing to nearlly 100%. Recommendations emphasize using a limited array of antibodies, eg. targets such as TTF-1, cytokeratin 5/6, and synaptophysin. In the infrequent case in which this does not make clear the cell type, a diagnosis of NSCLC, NOS is preferred over large cell carcinoma. Once a tumor is interpreted as adenocarcinoma, how well do we do in determining the predominant histologic subtype from the 2011 classification? The answer is poorly. This is related to both the small sample size with the recognition of the histologic heterogeneity within a sizeable tumor mass and the concept that the rather uniform manners in which neoplastic cells aggregate and present themselves in aspiration and exfoliative smears is typical of all types of adenocarcinomas and thus not representative of the histologic subtype. Current data supports the general notion that the predominant histologic subtype correlates with prognosis, and thus may serve as a morphologic grading surrogate. As just stated, it does not appear that cytology will permit such a parallel assessment. However, there is some evidence that certain nuclear attributes of adenocarcinoma cells in cytologic specimens are associated with prognosis and, hence, nuclear grading may be of value in this regard. Features which include nuclear contour, chromatin pattern, and the prominence of nucleoli can be used to formulate a meaningful nuclear grading system. This is likely better performed with alcohol-fixed Papanicolaou stained specimnes compared to air-dried Romanowsky stained samples. Note that mitotic figure counting is not a component of this proposal. It is crucial to recognize that cytologic samples provide a substrate for the molecular testing of therapeutically important mutations in adenocarcinoma cells which seem to be equal to histologic specimens. Thus, it is very relevant that sufficient cytologic material be collected at the very time of sampling and that it be utlized in an extremely judicious manner.

      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.

    • +

      MS21.2 - Molecular Diagnosis in Cytology and its Place in the New Classification (ID 557)

      14:00 - 15:30  |  Author(s): S.A. O'Toole

      • Abstract
      • Presentation
      • Slides

      Abstract
      There has been very little improvement in outcome from lung cancer over the last two decades but the identification of actionable mutations and structural rearrangements in subsets of patients with lung adenocarcinoma holds hope for the near future, particularly if these agents successfully move into the adjuvant setting. Detection of key molecular targets is central to this new understanding of lung adenocarcinoma and targeted therapy but poses significant challenges for implementation into routine clinical diagnostics. The importance of molecular testing in lung adenocarcinoma has been emphasized not only in the updated classification of adenocarcinoma but also in the recently released molecular testing guidelines for selection of lung cancer patients for EGFR and ALK Tyrosine Kinase Inhibitors produced by the College of American Pathologists, the International Association for the Study of Lung Cancer and the Association for Molecular Pathology this year with a central recommendation that tissue should be prioritized for EGFR and ALK testing. Recent molecular and genomic studies of lung adenocarcinoma in particular has resulted in the identification of other low incidence, novel driver mutations including structural rearrangements in ROS1 and RET-KIF5B as well as recognition of point mutations in BRAF and HER2 among others. It is apparent that “profiling” lung cancers for a range of important and potentially treatable driver mutations may offer significant advantages such as cost effective, rapid identification of actionable changes and efficient triage for clinical trials of novel agents. However performing molecular testing in lung cancer can be challenging given the majority of patients present with inoperable disease. This means histological and molecular diagnosis is generally performed on small biopsies including cytological specimens often with very limited material for testing. Furthermore much of this tissue has undergone formalin fixation and paraffin processing with subsequent DNA cross-linkage and fragmentation. There are additional problems of contamination with non-malignant tissue elements including stroma and inflammatory cells. There are also time pressures with the need for rapid results with current recommendations for results to be available within 10 working to allow appropriate triage for therapy. It is important to direct testing to appropriate clinical groups likely to benefit. While there are strong demongraphic associations of actionable mutations in lung adenocarcinoma, including non-smoking status, younger age, female sex and Asian ethnicity, these criteria are insufficiently robust to exclude patients without these characteristics from testing. Current recommendations in limited specimens are that molecular testing for EGFR or ALK gene changes be primarily undertaken in adenocarcinoma or cases with a component of adenocarcinoma. Fortunately cytology is emerging as a robust method for classification of lung cancer and these specimens are increasingly utilized for mutation testing. Large cell or histologically undifferentiated carcinomas with features suggestive of adenocarcinoma differentiation eg TTF-1 expression are also suitable for molecular testing. Molecular testing of limited biopsies may also be considered in cases showing squamous or small cell histology guided by clinical features such as ethnic background and non-smoking status among others. There is good concordance between primary and metastatic sites for EGFR mutational status and specimens from either site are acceptable for testing with choice based on morphological assessment of optimal specimens for molecular testing There are a wide variety of molecular techniques available to assess for the presence of key driver mutations in lung adenocarcinoma, each with their own limitations and advantages, but there is no perfect technology that fulfills all clinical and laboratory needs, especially on the limited material usually available in lung cancer mutation testing. Virtually all techniques for EGFR testing depend on PCR amplification, which is a major issue where limited DNA template is present raising the possibility of both false negative (due to sensitivity issues) and false positive results (eg due to amplification of formalin artifacts). In our own practice we have found that standard methods for DNA quantification such as spectrophotometry significantly overestimates the amount of DNA available for testing in comparison to more specific methods such as DNA fluorometry or estimates of amplifiable DNA copy number. We currently perform routine diagnostic mutation testing via multigene mutation profiling using a commercial panel, Oncocarta v1.0, on the massARRAY Sequenom platform in combination with fragment analysis for EGFR exon 19 and 20 insertions and deletions. This allows simultaneous determination of key mutations in EGFR and KRAS status as well as identifying rare but potentially actionable changes in BRAF, PIK3CA and HER2. In parallel, we perform immunohistochemistry for ALK and ROS1 to allow rapid triage for FISH testing if mutation profiling is negative. However not all cases have sufficient material for this approach and we are currently validating a new custom panel which can be performed reliably with less DNA. While cytological specimens are problematic in the generally small amount of tumour material available for testing, they have the advantage of often containing a relatively pure population of tumour cells, with a marked reduction in stromal contamination especially in FNAB specimen. Earlier studies suggested cytological specimen were less preferred to small tissue biopsies but a number of more recent publications have highlighted the suitability of these specimens. While the most recent recommendations suggest that cell block specimens allowing pretest morphological assessment are preferred to fresh smears, fresh material offers generally better quality and quantity of DNA. FISH testing for ALK gene rearrangement on cytological specimens is feasible and increasingly widely performed. Generally FISH testing requires less cellular material than mutation testing and the direct visualization of the molecular assay in tumor cells gives greater confidence that a negative result is far less likely to reflect problems with sensitivity as for EGFR testing. Expert morphological assessment is critical to ensure malignant cells are being assessed in this setting. In summary, cytology specimens are a commonly tested lung cancer diagnostic specimen and offer a number of advantages over more invasive small biopsies. Expert cytological assessment of specimens should be undertaken prior to molecular testing to maximize the quality and accuracy of testing.

      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.

    • +

      MS21.3 - Immunohistochemistry and the New Classification (ID 558)

      14:00 - 15:30  |  Author(s): G. Pelosi, A. Fabbri, M. Bimbatti, G. Leone, M. Garassino, F. De Braud, G. Sozzi, E.R. Haspinger, U. Pastorino

      • Abstract
      • Presentation
      • Slides

      Abstract
      Most lung cancers are readily diagnosed by using light microscope without attaining special stains [1], but at least 30% of NSCLC could benefit from immunohistochemistry (IHC) to unveil the cell differentiation lineages, especially when dealing with cytology and biopsy specimens [2]. Molecular methods, including micro-RNA expression analysis [3], are cumbersome and unlikely to be directly transferred into the everyday diagnostic workflow [2, 4]. IHC is not a perfect mathematic model, since there is a small (<5%) subset of NSCLC with ambiguous co-expression of glandular and squamous cell differentiation markers or negative reaction for any marker [5, 6]. a) What is the best combination of biomarkers to use? The coordinated expression for TTF-1 for adenocarcinoma and p40 for squamous carcinoma is currently emerging as the most reasonable and reliable biomarker duet in terms of sensitivity and specificity [2, 7-9]. Other promising biomarkers include napsin A for adenocarcinoma [10] and desmocollin-3 for squamous carcinoma [11, 12]. While TTF1 is the best marker for adenocarcinoma and p40 equivalent to p63 for squamous carcinoma, p40 is by far superior in terms of specificity since only rare adenocarcinomas are focally positive in comparison with p63 [2, 7-9]. b) Be aware of antibody clones and other technical issues. The monoclonal antibody 8G7G3/1 for TTF1 seems to be more specific for adenocarcinoma than other clones (such as SPT24) [8, 13], but it has also been recorded in gynecologic [14] and breast [15] carcinomas. Monoclonal antibody to napsin A for adenocarcinoma is less sensitive, but more specific than polyclonal antiserum [16]. The single best marker for squamous carcinoma is a polyclonal rabbit antiserum against p40 [7-9][,17], but very recently a monoclonal antibody has been made commercially available. c) Practical hints to surgical pathologists. NSCLC-NOS upon morphology with negativity for p40 and some TTF-1 positivity should be equated to poorly differentiated adenocarcinomas, once large cell neuroendocrine carcinoma (LCNEC) by relevant markers (e.g., synaptophysin) has been excluded. NSCLC-NOS upon morphology showing double negativity or with only erratic labeling for p40 in < 5% tumor cells in absence of TTF-1 should be considered as poorly differentiated non-squamous carcinomas corresponding, in most instances, to poorly differentiated adenocarcinoma once metastatic cancer has been reasonably excluded, keeping in mind however that the same immunoprofile may be shared by sarcomatoid carcinomas (excludible by morphology and vimentin IHC) [17] and LCNEC (excludible by synaptophysin IHC). Poorly differentiated squamous carcinomas are instead highlighted by strong and diffuse p40 expression and TTF-1 negativity, hence lack of p40 exclude by definition this tumor according to the axiom “no p40, no squamous” [9]. When morphology fails to conclusively subtype NSCLC, it is recommended specifying in the pathology report the real contribution of IHC to render the final diagnosis according to the relevant cell differentiation lineages (e.g., NSCLC-NOS, favor adenocarcinoma or squamous carcinoma by IHC) [6]. References 1. Travis W, Brambilla E, Muller-Hermelink H, Harris C. Tumours of the lung, pleura, thymus and heart. Lyon: IARC Press; 2004. 2. Rossi G, Pelosi G, Barbareschi M, et al. Subtyping non-small cell lung cancer: relevant issues and operative recommendations for the best pathology practice. Int J Surg Pathol 2013;21:326-36. 3. Lebanony D, Benjamin H, Gilad S, et al. Diagnostic assay based on hsa-miR-205 expression distinguishes squamous from nonsquamous non-small-cell lung carcinoma. J Clin Oncol 2009;27:2030-7. 4. Rossi G, Papotti M, Barbareschi M, Graziano P, Pelosi G. Morphology and a limited number of immunohistochemical markers may efficiently subtype non-small-cell lung cancer. J Clin Oncol 2009;27:e141-2; author reply e3-4. 5. Travis W, 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. J Thorac Oncol 2011;6:244-85. 6. Travis WD, Brambilla E, Noguchi M, et al. Diagnosis of lung cancer in small biopsies and cytology: implications of the 2011 International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification. Arch Pathol Lab Med 2013;137:668-84. 7. Bishop JA, Teruya-Feldstein J, Westra WH, et al. p40 (DeltaNp63) is superior to p63 for the diagnosis of pulmonary squamous cell carcinoma. Mod Pathol 2012;25:405-15. 8. Pelosi G, Fabbri A, Bianchi F, et al. DeltaNp63 (p40) and thyroid transcription factor-1 immunoreactivity on small biopsies or cellblocks for typing non-small cell lung cancer: a novel two-hit, sparing-material approach. J Thorac Oncol 2012;7:281-90. 9. Pelosi G, Rossi G, Cavazza A, et al. DeltaNp63 (p40) distribution inside lung cancer: a driver biomarker approach to tumor characterization. Int J Surg Pathol 2013;21:229-39. 10. Turner BM, Cagle PT, Sainz IM, et al. Napsin A, a new marker for lung adenocarcinoma, is complementary and more sensitive and specific than thyroid transcription factor 1 in the differential diagnosis of primary pulmonary carcinoma: evaluation of 1674 cases by tissue microarray. Arch Pathol Lab Med 2012;136:163-71. 11. Monica V, Ceppi P, Righi L, et al. Desmocollin-3: a new marker of squamous differentiation in undifferentiated large-cell carcinoma of the lung. Mod Pathol 2009;22:709-17. 12. Righi L, Graziano P, Fornari A, et al. Immunohistochemical subtyping of nonsmall cell lung cancer not otherwise specified in fine-needle aspiration cytology: a retrospective study of 103 cases with surgical correlation. Cancer 2011;117:3416-23. 13. Rekhtman N, Ang DC, Sima CS, Travis WD, Moreira AL. Immunohistochemical algorithm for differentiation of lung adenocarcinoma and squamous cell carcinoma based on large series of whole-tissue sections with validation in small specimens. Mod Pathol 2011;24:1348-59. 14. Siami K, McCluggage WG, Ordonez NG, et al. Thyroid transcription factor-1 expression in endometrial and endocervical adenocarcinomas. Am J Surg Pathol 2007;31:1759-63. 15. Robens J, Goldstein L, Gown AM, Schnitt SJ. Thyroid transcription factor-1 expression in breast carcinomas. Am J Surg Pathol 2010;34:1881-5. 16. Bishop JA, Sharma R, Illei PB. Napsin A and thyroid transcription factor-1 expression in carcinomas of the lung, breast, pancreas, colon, kidney, thyroid, and malignant mesothelioma. Hum Pathol 2010;41:20-5. 17. Pelosi G, Melotti F, Cavazza A, et al. A modified vimentin histological score helps recognize pulmonary sarcomatoid carcinoma in small biopsy samples. Anticancer Res 2012;32:1463-73.

      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.

    • +

      MS21.4 - AIS and the Well Differentiated Spectrum (ID 559)

      14:00 - 15:30  |  Author(s): Y. Yatabe

      • Abstract
      • Presentation
      • Slides

      Abstract
      According to the recent trend of increased frequency of CT screening for lung cancer, earlier stage of lung cancer is being detected and removed surgically. In applying the new classification to such lesions, it always becomes problematic to make a differential diagnosis among the three categories: adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA) and lepidic predominant adenocarcinoma. AIS is defined by a pure lepidic growth pattern with a continuous growth of neoplastic cells along the alveolar septa without disruption of the alveolar structures. MIA is defined as an adenocarcinoma with predominant lepidic growth with less than or equal to 0.5 cm area of sromal invasion. The definitions are summarized in Table. When the invasive area is more than 0.5 cm and lepidic growth is predominant, the tumor is diagnosed as lepidic predominant adenocarcinoma. The distinction is clinically important because AIS and MIA have been shown to have 100%5 year recurrence free-survival, whereas lepidic predominant adenocarcinoma can recur. However, the major difficulty for the differential diagnosis has in roots in the identification of stromal invasion and measurement of the invasive area. Histological stromal invasion is determined by tumor cell and stromal factors. Because the area where the tumor cells show invasive structure is regarded as an invasive area, it is important to recognize the differentiation of lepidic pattern from papillary or acinic pattern. However, the distinction is often difficult in practice. In terms of the stromal factor, it is also difficult to differentiate an invasive scar (myofibroblastic stroma) from a scar due to collapse. Physicians should know this room for discussion, and practical solutions should be shared with pathologists. Figure 1

      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.

    • +

      MS21.5 - Molecular Analysis for Distinction Second Primary of the Lung vs Lung Metastasis? (ID 560)

      14:00 - 15:30  |  Author(s): S. Dacic

      • Abstract
      • Presentation
      • Slides

      Abstract
      Background: The development of high-resolution chest imaging techniques and screening of smokers for lung cancer resulted in increased detection of multiple lung cancers. The challenge for pathologists and treating physicians is to determine whether multiple lung cancers represent separate primary tumors or metastasis, as this affects the stage, treatment and prognosis. The clinical and pathological criteria used to define multiple lung tumors were initially published by Martini and Melamed in 1975. These criteria are based on tumor morphology and location and may not predict prognosis. The AJCC 7[th] edition staging of lung adenocarcinoma recognized shortcomings of this proposal and incorporated changes in the staging of multiple lung cancer, but molecular genetic analysis was not recommended as a standard approach. Methods: PubMed available peer-reviewed original articles and experience of the author. Results: Distinction between primary tumors and intrapulmonary metastasis becomes challenging when tumors are morphologically similar. Since the original Martini and Melamed proposal, many molecular approaches have been utilized in the evaluation of clonal relationships between multiple lung nodules including DNA microsatellite analysis, PCR assays for common somatic mutations, aCHG, and gene expression analysis. Molecular classification of multiple lung cancers is concordant with pathological classification in about two-thirds of the cases. It is difficult to determine the precise percentage because of the relatively small number of analyzed cases, mixed analysis of synchronous and metachronous tumors, and use of different methods and interpretation criteria. Early studies used two types of clonality assays: a panel of variable number of polymorphic microsatellite markers and X-chromosome inactivation analysis (Am J Surg Pathol 2005; 29(7):897;Ann Diagn Pathol 2001;5(6):321; Clin Cancer Res 2000; 6(10):3994; J Natl Cancer Inst 2009;101:560) . Tumors with largely concordant results were considered clonal in origin (metastases), and those with discordant findings were considered to be independent primary tumors. The main weakness of earlier studies was a limited number of analyzed genes. Recently, more comprehensive approaches analyzing a large number of single nucleotide polymorphic loci in a single assay or large-scale DNA sequencing of tumors were used (Clin Cancer Res 2009; 15(16):5184; Lung Cancer 2012;77:281). Although more comprehensive molecular approaches were used, a proportion of cases with discordant molecular and morphological results remained similar. Furthermore, molecular profiling only slightly improved prognostic classification of multiple lung tumors. Standard practice is to test non-resectable adenocarcinomas for common actionable somatic mutations (e.g. EGFR) and gene rearrangements (e.g. ALK) as predictors of response to targeted therapies. This information can also be used for improved staging of multiple lung nodules (Eur Resp J 2012; 39:1437; Lung Cancer 2012; 77:281). Based on similar or different mutational profile synchronous tumors may be classified as independent primaries or intrapulmonary metastases. It is very likely that surgically non-resectable tumors with different mutational profiles such as EGFR and KRAS will show different treatment responses, further emphasizing the need for separate analysis of multifocal tumors. In contrast to morphologic classification, molecular profiling can be performed on the cytology specimens. This approach can be used in adenocarcinoma only, and currently no standard molecular testing in squamous cell carcinoma is in practice. Conclusions: Molecular approaches to classification of multiple lung tumors have not been standardized, and their performance in routine clinical practice remain to be established. Testing for common activating oncogenic mutations and translocations is likely to provide information about clonal relationship between multifocal lung tumors. Implementation of molecular information to current histologic staging could improve the accuracy of staging in patients with multifocal tumors and improve therapeutic decision making.

      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

  • +

    MO26 - Anatomical Pathology II (ID 129)

    • Event: WCLC 2013
    • Type: Mini Oral Abstract Session
    • Track: Pathology
    • Presentations: 1
    • +

      MO26.14 - Histological Prognostic Markers in Squamous Cell Carcinoma of the Lung (ID 2908)

      10:30 - 12:00  |  Author(s): A.L. Moreira

      • Abstract
      • Presentation
      • Slides

      Background
      The current IASLC/ERS/ATS classification of pulmonary adenocarcinoma indicates that different patterns of growth in adenocarcinoma are associated with prognostic value. There is however, very little information concerning histological prognostic markers in squamous cell carcinomas of the lung. In contrast to adenocarcinoma, squamous cell carcinoma is more homogeneous histologically. However, the World Health Organization classification of lung tumors recognizes different patterns of growth in squamous cell carcinomas. In this study we evaluated several histological parameters including growth patterns and nuclear features and their association with prognosis in a population of stage 1 squamous cell carcinomas.

      Methods
      A cohort of 165 stage I squamous cell carcinomas of the lung were evaluated. The presence of different histological growth patterns such as papillary, infiltrative, pushing borders, intraalveolar, pseudo-glandular, basaloid, small nest and presence of infiltrating single cells, as well as the cell type (clear cell, transitional, syncytial, and glassy) were evaluated in a semi- quantitative manner by recording the percent of each histological pattern or cell type with 10% increments totaling 100% for tumor. In addition, the presence of peripheral palisading, nuclear features (nuclei size, chromatin patterns, nuclear contour, presence of nucleoli, and mitotic figures), and keratinization were also evaluated. The association of predominant pattern of growth, cell type, and nuclear features with recurrence free survival (RFS), characterized by time to recurrence or death of disease and overall survival (OS) were evaluated.

      Results
      There were 66 women and 97 men in this population with a mean age of 75±9 year old. All patients were smokers. The mean follow-up was of 47.8 months (4 years). Among histological growth patterns, tumors with predominant papillary and pushing borders appear to have a slightly better outcome compared to other predominant patterns of growth (RFS p=0.05 and OS 0.025). It is interesting to note that squamous cell carcinomas with a predominant basaloid growth pattern, which is considered to be a pattern of poor differentiation, did not have worse prognosis copared to other features. There was no association of cell type, nuclear features, presence of palisading or keratinization with prognosis. There was no difference of nuclear features among tumors with different growth patterns and cell types.

      Conclusion
      Squamous cell carcinomas appear to be more homogeneous than adenocarcinomas of the lung despite some histological variances. Evaluation of several histological parameters like growth pattern, cell type, and nuclear features failed to indicate a strong association of any of these parameters with prognosis, with exception of papillary and pushing border growth patterns that when present as predominant patterns of growth were associated with a better prognosis. This suggests that contrary to adenocarcinoma, a histological based grading system may not be easily established for squamous cell carcinomas of the lung.

      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.

  • +

    O17 - Anatomical Pathology I (ID 128)

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

      O17.05 - Accuracy and Interobserver Agreement in Identifying Histologic Subtypes in Stage I Lung Adenocarcinomas ≤3 cm Using Frozen Section (ID 2590)

      10:30 - 12:00  |  Author(s): A.L. Moreira

      • Abstract
      • Presentation
      • Slides

      Background
      The new IASLC/ATS/ERS classification of lung adenocarcinoma (ADC) histologic subtypes is now recommended for prognostic stratification. The ability to determine histologic subtype accurately by frozen section (FS) may help surgeons to choose limited resection versus anatomic resection in the management of lung ADC. The aim of this study is to investigate the accuracy and interobserver agreement of FS for predicting histologic subtype.

      Methods
      FS and permanent section slides from 361 surgically resected stage I lung ADCs ≤3 cm were reviewed for predominant histologic subtype and presence or absence of lepidic, acinar, papillary, micropapillary, and solid patterns. To determine interobserver agreement, 50 cases were additionally reviewed by 3 pathologists. To test the accuracy of FS in determining degree of invasion in cases with predominantly lepidic growth pattern, 5 pathologists reviewed FS slides from 35 patients and attempted to discriminate between adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and lepidic predominant adenocarcinoma (LPA).

      Results

      Parameter Accuracy, % (95% CI) Sensitivity, % (95% CI) Specificity, % (95% CI) κ
      Predominant histologic subtype
      Overall 68 (63–73) Not applicable Not applicable 0.565
      Lepidic 90 (86–92) 75 (64–84) 93 (90–96) 0.681
      Acinar 76 (71-80) 70 (61–77) 79 (73–84) 0.481
      Papillary 85 (81-88) 62 (50–72) 91 (87–94) 0.527
      Micropapillary 94 (91-96) 21 (9–40) 99 (97–100) 0.277
      Solid 91 (88-94) 79 (67–87) 94 (90–96) 0.700
      Presence or absence of each histologic pattern
      Lepidic 80 (76–84) 75 (69–80) 91 (84–96) 0.588
      Acinar 89 (85–92) 90 (86–93) 67 (35–90) 0.252
      Papillary 72 (67–77) 70 (64–75) 79 (69–87) 0.397
      Micropapillary 67 (62–72) 37 (30–45) 94 (89–97) 0.321
      Solid 84 (80–88) 69 (61–76) 96 (92–98) 0.670
      The accuracy of FS for predicting histologic subtype is shown in the Table. There was moderate agreement on the predominant histologic subtype between FS diagnosis and final diagnosis (κ=0.565). FS had high specificity for micropapillary and solid patterns (94% and 96%, respectively), but sensitivity was low (37% and 69%, respectively). The interobserver agreement was satisfactory (κ > 0.6, except for acinar pattern). All cases of AIS were correctly diagnosed using FS. For MIA, only 41.3% of FS diagnoses were correct, and 52% were overdiagnosed as LPA; for cases of LPA, 79% of FS diagnoses were correct.

      Conclusion
      FS can provide information on the presence of aggressive histologic patterns—micropapillary and solid—with high specificity but low sensitivity. FS is not suitable for determining the predominant pattern or degree of invasion. Although FS can be helpful in diagnosing AIS, it has poor accuracy in distinguishing MIA from LPA.

      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.

  • +

    P2.15 - Poster Session 2 - Thymoma (ID 191)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Thymoma & Other Thoracic Malignancies
    • Presentations: 1
    • +

      P2.15-007 - Digital Microscopy Reproducibility Study of Thymic Epithelial Neoplasms (ID 2883)

      09:30 - 16:30  |  Author(s): A.L. Moreira

      • Abstract

      Background
      Thymic epithelial tumors are rare and morphologically heterogeneous which constitutes interpretive challenges to practicing pathologists. Advances in digital imaging provide an opportunity to disseminate knowledge of these rare tumors, and can be potentially useful as diagnostic and educational tools. However the diagnostic reproducibility utilizing digital slide imaging needs to be validated.

      Methods
      Twenty cases of thymomas or thymic carcinomas with characteristic morphologic features were scanned into the APERIO system. The images were sent to pathologists with expertise in thoracic pathology in 6 different centers. The pathologists were asked to classify the tumors according to the World Health Organization (WHO) 2004 classification and to evaluate invasion on the scanned material. In addition, they were asked to indicate their confidence in the diagnosis using the imaging system. Interobserver agreement was evaluated. After discussions of the first 20 cases, a second round representing 10 cases were evaluated by digital images by the participating pathologists.

      Results
      In the initial phase, there was agreement among pathologists for the diagnosis of thymoma and thymic carcinoma in 75 % of cases (n= 14), in the remaining 6 cases, the disagreement was between cases of B3 thymoma and thymic carcinoma in five and between Type A thymoma and thymic carcinoma in one (kappa=0.43, moderate agreement). Perfect agreement was seen in 4 thymoma cases, where all pathologists diagnosed the same WHO type. These were classical cases with pushing borders and large fibrous bands. In other cases there were disagreements among the classification of the tumor as B2, B3, and AB. The cases with most disagreement were histologically heterogeneous with combined patterns. When invasion was evaluated, the overall k coefficient is 0.49 for the presence of invasion. In the second round of cases, we observed an improvement in interobserver agreement for diagnosis thymoma vs thymic carcinoma (kappa = 0.63) and for determination of invasion (present versus absent) (k=0.57). Most pathologists found that the digital images were comparable with glass slides and the overall confidence in the diagnosis was good.

      Conclusion
      The diagnostic accuracy of thymic epithelial tumors by digital images is equivalent to that reported in prior studies using glass slides. Digital imaging is a good tool for remote consultation and educational purposes. In the majority of specimens, pathologists are able to make the correct diagnosis. Major challenges include distinguishing B3 tumors and carcinomas and tumors with morphologic heterogeneity. The overall agreement can be improved after training. This technology could be used to establish a digital slide bank which could provide a method for training pathologists with less experience in the pathology of thymic epithelial tumors, to foster collaborative work in the field, and diagnostic consultation.