Maximizing the yield of lymph node cytology

Lessons learned from rapid onsite evaluation of image- and endoscopic-guided biopsies of hilar and mediastinal lymph nodes

Abstract

The evaluation of mediastinal and hilar lymph nodes for tissue diagnosis and staging of lung cancer is now commonly performed by minimally invasive, nonsurgical procedures such as computed tomography-guided fine-needle aspiration and endobronchial ultrasound-guided transbronchial needle aspiration. Ensuring that a sufficient quantity of cellular material has been acquired to enable multiple studies has become a priority issue in the era of personalized medicine, especially for patients with lung cancer, and this can be accomplished by rapid onsite evaluation (ROSE). This commentary focuses on the use of ROSE in guided procedures, especially for hilar and mediastinal lymph node aspirates, and describes an algorithm for handling these specimens. Cancer (Cancer Cytopathol) 2011;. © 2011 American Cancer Society.

Rapid advances in oncologic therapies have mandated the assessment of molecular markers during the diagnostic workup of small tissue samples as the standard of care. For example, in patients with lung cancer, several studies have established epidermal growth factor receptor (EGFR)-activating mutations as the key biomarkers in selecting patients for first-line therapy with EGFR tyrosine kinase inhibitors.1 Accordingly, ensuring that a sufficient quantity of viable tumor has been obtained from individual patient lesions to enable multiple studies has become a priority issue.

The evaluation of mediastinal and hilar lymph nodes for tissue diagnosis and staging of lung cancer is now commonly accomplished by minimally invasive, nonsurgical procedures such as computed tomography-guided fine-needle aspiration (CT-guided FNA) and endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA). In some centers, these techniques have been incorporated into routine clinical practice, largely replacing surgical interventions for the procurement of diagnostic samples, especially for patients who are considered ineligible for curative surgery. Although the primary objective of cytopathology continues to be the provision of a precise diagnostic evaluation, the current requirement for pathologic subtyping of lung cancer and the evaluation of prognostic markers for patient treatment have imposed new challenges on cytopathologists.

By providing immediate specimen evaluation for adequacy and through the use of novel specimen preservation techniques, cytopathologists play a critical role in aliquoting precious samples and ensuring that the relatively small quantity of cells often available from those samples can provide sufficient reliable molecular data.2 In a recent consensus article concerning EGFR mutation testing in patients with non-small cell lung cancer (NSCLC), the necessity of standardization for tumor specimen handling, including fixation, as a means of ensuring better preservation of the DNA structure and guaranteeing proper mutation analysis was emphasized.3

Based on our previous experience with lymph node aspirates,4, 5 the study institution has developed an algorithm for the rapid onsite evaluation (ROSE) of mediastinal lymph nodes that is designed to extract all relevant diagnostic and prognostic information from minimal cytologic material (Fig. 1). Our main objective is maximizing cell recovery to meet the current demand for the use of multiple ancillary techniques and molecular analyses in thoracic oncology. The current commentary focuses on our process for triaging and handling lymph node aspirates obtained by CT-guided and EBUS-TBNA procedures. We also discuss new and alternative approaches for storing FNA specimens.

Figure 1. Algorithm for handling lymph node samples is shown. For rapid onsite evaluation (ROSE), air-dried smears are prepared and stained using the Romanowsky method. Smears are also fixed in a Carnoy-ethanol solution for Papanicolaou staining. Needles are rinsed in saline. Suspensions from saline-rinsed needles are submitted for cell count. If only lymphoid cells are present, the specimen is sent for immunophenotyping by flow cytometry (FC) or laser scanning cytometry (LSC). If enough cells are obtained, cytospin preparations for fluorescence in situ hybridization (FISH) studies and for MIB-1 proliferation index by immunohistochemistry (IHC) are generated. If carcinoma is suspected and fragments are present in the cell suspension, they are used for cell blocks and ThinPrep slides are produced. If an infectious process is suspected, an aliquot is sent to microbiology. An aliquot can also be used for other alternative strategies for cellular preservation such as cryopreservation and FTA cards. FFPE indicates formalin-fixed, paraffin-embedded; FITC, fluorescein isothiocyanate; TTF-1, thyroid transcription factor; EBER, Epstein-Barr virus-encoded RNA; EGFR, epidermal growth factor receptor.

Imaging- and Endoscopic-Guided Procedures

At the study institution, the most common nonsurgical procedures used for obtaining a tissue diagnosis from hilar and mediastinal lymph nodes or masses are CT-guided mediastinal FNA and EBUS-TBNA.

CT-guided mediastinal FNA

CT-guided mediastinal FNAs are performed by radiologists with subspecialty interest and training in thoracic and interventional radiology or the radiology fellows under their supervision.

EBUS-TBNA

EBUS-TBNA is performed by physicians with training in bronchoscopy, who often are pulmonologists. In our institution, as in many others, thoracic surgeons perform all these procedures. The detailed procedure has been previously described.6

Smear Preparation and Specimen Handling

The aspirated material obtained from lymph nodes by CT-guided FNA or EBUS-TBNA is smeared onto glass slides for cytologic interpretation using the Löwhagen “1-step” technique.7 A single drop of material is placed on the slide and in most cases it can be divided by the “touch-off” technique so that 1 to 3 slides can be air-dried and at least 1 can be fixed in an acetic acid-ethanol (modified Carnoy's) solution. The “touch off” technique for splitting material to produce multiple smears is comprised of lowering a clean slide onto the previously expelled drop and picking up a small amount of the sample to make additional smears. The air-dried smears are stained with Romanowsky stains (Giemsa stain or Field stain) and the ethanol-fixed slides are stained by the Papanicolaou technique. The needles are then rinsed in sterile saline to allow recovery of the maximum number of cells for the necessary ancillary studies. We use commercial disposable containers to keep saline sterile until its use. Saline is preferred because of its low cost and low interference with flow cytometry, immunohistochemistry, and molecular and microbiology studies.

ROSE

ROSE, performed by a cytopathologist after sampling at each lymph node location, is used to assess the quantity of cellular material and enable a preliminary diagnosis that will determine optimal handling of the needle rinse. In other institutions, cytotechnologists are an alternative to cytopathologist attendance.8

Similar to other authors,9, 10 we regard the macroscopic examination of the sample before staining to be an important parameter, knowing that an experienced cytopathologist can derive considerable information from visual inspection of the sample and the smears made from it. Although to the best of our knowledge standard criteria that are predictive of microscopic adequacy have not been established through systematic assessment of the macroscopic characteristics of aspirates, a simple macroscopic grading system has been shown to be useful when pathologists are not available.11 The characteristics evaluated are generally the consistency of the sample and the presence and amount of blood, granular material, small tissue fragments, pus, and necrotic material. These features are noted before, during, and after smear preparation. In the case of EBUS-TBNA samples, we also note the amount of airway mucus and the presence of cartilage fragments. A cellular sample containing diagnostic material is often thick and pale, with a consistency similar to cream and smears demonstrate a glistening granular appearance in which small tissue fragments are evident.9 This granularity is more evident in cases of carcinoma. Material from a cellular lymphoid lesion yields a uniform and opaque smooth rather than granular smear, and a necrotic sample often appears “greasy” and may demonstrate parallel lines of flocculated material. An overtly enlarged, suspicious lymph node that yields no material despite a determined FNA attempt is suggestive of nodular sclerosing Hodgkin lymphoma or some other fibrotic process such as a mediastinal large B-cell lymphoma.9

After microscopic examination, the cytopathologist immediately reports the findings and notifies the operator when sufficient material has been obtained. Samples comprised of only mucus, a preponderance of bronchial or cartilage cells, scant or no lymphocytes, necrotic material, and no specifically diagnostic material are considered inadequate. Additional passes are performed in such cases, provided patient tolerance permits it and no complications develop.

The specimens are triaged to appropriate ancillary studies according to the preliminary cytomorphologic diagnosis. If infection is suspected and/or slides demonstrate granulomas, with or without a necrotic background, an aliquot from the needle rinse is sent for microbiological analyses, including cultures for tuberculosis and fungi or Gomori methenamine silver and Ziehl-Neelsen stains. The local microbiology laboratory should be consulted regarding their preferred transport medium.

If carcinoma is considered in the preliminary cytomorphological assessment, the needle rinse is processed to produce ThinPrep slides and/or formalin-fixed, paraffin-embedded (FFPE) cell blocks (CBs).

If large cell lymphoma or small lymphoid proliferations with a clinical and morphologic differential diagnosis of reactive lymphadenopathy and indolent lymphoma are considered, the sample is submitted for cell surface marker immunophenotyping. When lymphoma is confirmed by immunophenotyping, the proliferation index is assessed by immunohistochemistry for MIB-1.

Ancillary techniques

All ancillary studies are performed using the cell suspensions from the needle rinses. Blood clots and visible tissue fragments are removed for preparation of a CB and remaining cells in suspension are used for cytospin preparations (CPs), with the number of slides being determined by the cell count in the suspension (disposable Neubauer counting chamber). CPs are produced (usually 4-6 slides) and used for fluorescence in situ hybridization (FISH) studies and MIB-1 immunohistochemistry.

Immunophenotypic Profiling

Analyses are performed on fresh cell suspensions from needle rinses using laser scanning cytometry (LSC) and/or flow cytometry. RPMI is an alternative medium for collecting cells for flow cytometry. However, if RPMI is used for culture it is important to ensure that it has not been augmented with antibiotics. We have not tested other alternatives such as phosphate-buffered saline. We are also unaware of the results of producing CBs with cytologic material collected in RPMI. A minimum total of 15,000 cells are required for a full immunophenotyping panel by LSC. The complete panel that is usually assessed is comprised of the following monoclonal antibodies: CD19, CD20, CD3, CD5, CD10, kappa, lambda, CD22, CD23, FMC7, CD4, CD8, CD56, CD7, CD2, CD25, CD14, CD11c, and CD16. The antibody clones and concentrations, fluorochromes, and suppliers have been previously detailed.4 A limited panel tailored to the morphologic differential diagnosis is performed in cases with low cellularity. Interpretation of immunophenotyping data is performed by the cytopathologists and integrated with the cytomorphological findings in the final cytologic report.

MIB-1 Proliferation Index and FISH Assays

MIB-1 immunocytochemistry is performed on CPs fixed in 95% ethanol. Details regarding the protocol, the antibody, and the method of evaluation for positive cases have been described in a previous publication.12

FISH assays are performed on CPs according to the protocol described previously.5 CPs that are created by preserving whole intact nuclei provide a good alternative to CB sections and avoid the nuclear truncation artifacts inherent in the sectioning of FFPE CBs. In addition, using CPs for FISH studies can save all possible FFPE CB sections for immunohistochemistry. One or more of the following probes is used: IGH/BCL2 to detect t(14;18) (q32;q21.3), IGH/CCND1 to detect t(11;14)(q13;q32), IGH/MYC:CEP 8 to detect t(8;14)(q24;q32) (dual-color, dual-fusion probes), IGH, mucosa-associated lymphoid tissue 1 (MALT1), and MYC (break-apart probes) as well as enumeration/copy number probes CEP 3 (D3Z1), CEP 12 (D12Z3), 13q14 (D13S319), 13q34 (D13S25/LAMP1), ATM/D11Z1, 17p13 (TP53), and CEP 17 (D17Z1) (Abbott/Vysis, Des Plaines, Ill). The selection of probes is based on initial clinical, cytomorphologic, and immunophenotyping data.

CB Preparations

FFPE CBs prepared according to the HistoGel method (Thermo Fisher Scientific Inc “Richard Allan Scientific”, Kalamazoo, MI) are preferentially used for immunohistochemistry studies for histologic subtyping or to determine a primary tumor site. They may also be used for EGFR mutation analysis in nonsquamous lung carcinoma, and Epstein-Barr virus-encoded RNA in situ hybridization (EBER-ISH) to provide supporting evidence for some lymphoma subtypes and undifferentiated carcinoma of nasopharyngeal type.

Immunohistochemistry for the identification of Reed-Sternberg cells can also be performed on CB sections using monoclonal antibodies against CD15, CD30, CD45, and CD20.

EGFR Mutation Analysis

Mutation analysis is performed with DNA extracted from 5 tissue sections measuring 7 μm in thickness and taken from FFPE CB sections.

Other Alternatives and Novel Specimen Preservation Techniques

New molecular assays require high-quality DNA and therefore are sensitive to initial specimen handling and DNA integrity. Recent studies have demonstrated encouraging results with the use of cytological samples.

Our group has described a new simple, rapid, and effective procedure for the collection and storage of cytologic material for mutational analysis using DNA extracted from FTA cards that yields better quality DNA for polymerase chain reaction than that extracted from FFPE CBs.13 Furthermore, a recent study has demonstrated the preservation and recovery of higher molecular weight DNA from archival FNA smears, which was used for high-resolution comparative genomic hybridization array, DNA methylation, and single nucleotide polymorphism genotyping platforms.14

The use of fresh cells whenever possible may be preferable because DNA will be less degraded. The storage of fresh cells relies mainly on cryopreservation techniques. It recently was demonstrated that the cellular morphology and RNA integrity necessary for studies of gene expression are maintained with cryopreserved samples.15

The detection of phosphorylated signaling proteins in FNAs for the measurement of pathway activity using multiplexed analysis has been described and presents the opportunity for its application in drug response studies and functional assays.16

New miniaturized devices using different platforms may add useful data to morphological assessment and could be integrated into the clinical routine to improve the speed with which results are determined.17

Conclusions

Minimally invasive procedures can provide sufficient samples for routine diagnostic workup, including required ancillary techniques, as well as good-quality and reliable material for molecular tests. ROSE by an experienced cytopathologist plays a critical role in ensuring that lymph node samples are handled properly and are appropriately divided for diagnostic purposes and, when patient consent permits, for future research projects. The wide use of cytological samples for multiple molecular platforms and for the evaluation of multiple biomarkers will require the active participation of cytopathologists in optimizing sample handling so that relevant diagnostic and prognostic information for treatment decisions is extracted and material remains for future studies.

FUNDING SUPPORT

No specific funding was disclosed.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.