Validation of immunocytochemistry as a morphomolecular technique
Abstract
Immunocytochemistry (ICC) is a long-established means for clinical laboratories to investigate material for which it is difficult to obtain tissue samples. Unlike immunohistochemistry (IHC), the cells do not retain surrounding tissue environment/architecture. This can be of benefit in that fixation is often immediate and rapid, protecting the cells. Although fixation is frequently observed as the main preanalytic variable of test quality, all cytology preanalytic factors should be identified and controlled. In addition, the validation of ICC should take the same rigorous approach that other molecular pathology techniques follow. A three-step validation protocol is offered here. The end result is a comprehensive, morphomolecular approach to ICC, with an emphasis on therapeutic ICC. Cancer Cytopathol 2016;124:540-5. © 2016 American Cancer Society.
ICC—VALIDATION VARIABLES IN THE MOLECULAR ERA
It is crucial to establish a basis for the validation of immunocytochemistry (ICC) as a method of managing uncertainty. We have previously published some recommendations to immunohistochemistry (IHC) testing, identifying the primary use of tests as diagnostic, genetic, or therapeutic.1 The latter is very important, because it is part of what has been named molecular diagnostic cytopathology.2, 3 Indeed, analyses of human epidermal growth factor receptor 2 (HER-2/neu) or anaplastic lymphoma kinase (ALK) (by ICC or in situ hybridization (ISH)), among others, must be regarded as bona fide molecular tests.
In the diagnostic setting, Sauter et al4 reported on their verification of ICC on alcohol-fixed cytologic material over a 5-year period comparing cytology findings with formalin-fixed, paraffin-embedded (FFPE) tissue samples. All 71 antibodies included in that study had undergone validation according to the College of American Pathologists (CAP).5 That work consisted predominantly of cell preparations from FNA samples of lung, lymph node, pleural cavity, and liver. Compared with tissue-FFPE results, the authors confirmed that the methodology used met the demands of the diagnostic service, although they identified 7 antibodies that were more suitable for tissue-FFPE material (cluster of differentiation 5 [CD5], a type I transmembrane protein; c-kit/CD117 [proto-oncogene c-Kit or tyrosine-protein kinase kit]; inhibin; napsin A; octamer-binding transcription factor 3/4 [OCT3/4]; and paired box 5 [PAX-5]). In our opinion, this work demonstrates a reality sometimes known in routine practice but difficult to substantiate, namely: 1) that not all antibodies are equally suitable for ICC and IHC; and 2) that ideal, fully validated standard operating procedures for tissue-based and cell-based immunochemistry using the same antibody may vary.
These considerations are particularly important for clinical scenarios in which only the availability of cytology samples can dictate the application of personalized medicine to the patient. In the therapeutic setting, Bueno Angela et al6 demonstrated that cell blocks prepared from FNA samples could be used successfully for the assessment of estrogen receptor (ER), progesterone receptor (PR), and HER-2/neu proteins by IHC compared with surgical specimens. These were of particular utility in obtaining samples from patients with inoperable and recurrent disease. In line with tissue-FFPE samples, the authors recommended that subsequent consideration be given for appropriate follow-up with fluorescence in situ hybridization analysis of low or equivocal Her2 results, although it should be acknowledged that more studies need to be done.
In our opinion, cytology-specific validations are at the core of the management of uncertainty for meeting accreditation standards like those promulgated of the International Organization for Standardization (ISO) 15189 (2012) and the CAP. This means that the various types of slide preparations available need to be considered. These include smears, cytospins, commercial liquid-based preparations and cell blocks from commercial kits, and noncommercial cell block methods, such as agar. Also, whereas cell blocks may be FFPE, cytospins may be fixed in alcohol. The implications of each variable for cell preparation need to be identified, assessed, and controlled and should be documented as part of the validation process.
There is a trend in clinical laboratories to perform ICC on FFPE cell blocks. Kirbis et al7 reported on 6-year returns to the UK National External Quality Assessment Service (UKNEQAS) ICC and ISH Cytology module. In this system, standardized cytospins had been circulated to participants according to the protocol of Maxwell et al,8 and participants were asked to stain for named antigens (CD3, CD45, calretinin, or cytokeratin) and return these along with their material used for quality control (in-house material). Only 34% of participants used cytology preparations that were not FFPE cell blocks for their in-house returns. The remaining 66% of the participating laboratories that made such returns used either FFPE tissue sections or FFPE cell blocks. UKNEQAS ICC and ISH have since introduced an option for participants to receive cell block material as the source material. Initially, 61% of participants requested this format in November 2014. By July 2015, this had risen to 71%.
With such variables in mind, the preanalytic landscape for cytology ICC can prove difficult to standardize across different laboratories; but, whatever system is in use within each laboratory, local knowledge of these parameters, validation, and rigorous quality control and audit are essential to the management of uncertainty.
ICC—WHEN CYTOLOGY IS EVERYTHING YOU'VE GOT
The current paradigm of cytology-based, personalized medicine is endobronchial ultrasound (EBUS)-guided transbronchial needle aspiration (EBUS-TBNA). This is a valuable technique for staging, diagnosis, and molecular testing in lung cancer9-11 and is often the only material available for diagnostic or therapeutic purposes. The opportunity may also arise for the collection of material for onsite assessment and/or the triaging of a specimen in the determination of its suitability for further studies, such as ICC. In the non-lymph node setting, Yang et al9 demonstrated that EBUS was not only a highly sensitive method in helping to distinguish benign from malignant lesions but also that ICC added value in the diagnosis of such lesions. Moreover, in a retrospective study conducted between 2005 and 2009, Sanz-Santos et al10 confirmed that EBUS-FNA can deliver high-quality samples from enlarged intrathoracic lymph nodes that are suitable for both diagnostic and molecular tests.
Because of the requirement for good-quality cytology samples combined with effective ICC panels, we need to consider molecular testing as part of a personalized, stratified medicine approach that has extra diagnostic relevance to challenging areas in cytopathology with high levels of diagnostic uncertainty, such as pancreatic FNAs or thyroid aspirates.12-17 Therefore, it is essential for cytology laboratories to capture and control preanalytic parameters and perform cytology-specific validations, hence minimizing uncertainty.
ICC—WHAT CONSTITUTES AN ADEQUATE VALIDATION?
We recently published an article on the measurement of uncertainty in tissue-based IHC, particularly in relation to ISO 15189 (2012) standards (Fig. 1).18 The validation of ICC requires standards of validation similar those for IHC. Our recommendations rely on sourcing and using well characterized control material. In cytology, this can be problematic. Excess material with the augmentation of well characterized cell lines from reputable suppliers may provide the bulk of daily quality-control material. To reflect the quality of the test material with a high level of confidence for the interpretation of results, it is essential that preanalytic parameters are duplicated or are reflected in both test and control material. With a trend toward the use of FFPE cell blocks, maintenance of controls is less of a problem than for other cytology preparations. The use of alcohol fixation, combined with the protection polyethylene glycol can offer, may be one solution for cytospin preparations.8
Primary antibody uncertainty: how do we validate? Modified from Elliott K, McQuaid S, Salto-Tellez M, Maxwell P. Immunohistochemistry should undergo robust validation equivalent to that of molecular diagnostics. J Clin Pathol. 2015;68:766-77018.
The level of experimentation required to fully characterize antibodies in steps 1 through 3 of our model (Figure 1 and below) may be beyond the scope of the majority of diagnostic laboratories but may be undertaken by more academic or industry-based laboratories. Smaller diagnostic laboratories, therefore, may only be able to verify the transfer of techniques rather than full validation.19
Step 1. What Is Known About the Antibody to be Validated?
Reference material is used to confirm the staining characteristics of the antibody as established by the manufacturer/supplier. At the first level of determining what is known about an antibody, well characterized antibodies are often used for diagnostic purposes for which the supplier/manufacturer has a detailed antibody data sheet. The second and third levels of antibody characterization will depend on testing undertaken by methods such as flow cytometry, Western blot analysis, and/or enzyme-linked immunosorbent assays and the use of knock down or transcription of the target.20, 21
Step 2. Is There a Good Protocol?
The workup of the technical aspects of antibodies uses control material in the format of the testing parameters. If the cytology laboratory offers more than 1 format, eg, cytospins and cell blocks, then the protocols need to be developed accordingly with any pretreatment, including an evaluation of antigen-retrieval options and optimization of dilutions. Antigen retrieval is required for formalin-fixed cells; however, when samples are fixed in alcohol or other nonformalin fixatives, it is unlikely to be needed. In the case of cell blocks that have been fixed and processed alongside clinical material, well designed tissue microarrays with cores from surgical specimens that represent a range of expression values from no expression through high expression can be used for this workup. Knowledge of cross-reactivity is essential for accuracy and precision. This is of particular importance for cases in which an antibody may be well characterized for tissue sections but immunoreactivity in whole cell preparations needs to be compared and contrasted. Antibodies that have prognostic significance but are infrequently encountered also fall within this category.
Step 3a: Is it Fit for Purpose?
This is a key step in immunochemistry (IHC or ICC) marker validation. This step follows the technical validation and informs the validity of adopting protocols and certain antibodies themselves in preference to others. For example, how is the antibody contributing to the clinical situation? Is it part of a larger diagnostic panel? If so, then the gold standard will be for a defined purpose, such as the level of concordance with reference material and how the results of the new antibody add value to the new diagnostic panel. Current CAP guidelines as a consensus expert opinion state that nontissue processed cytologic specimens should test sufficient numbers of cases to ensure consistency.5 For concordance analysis between the reference standards and testing of the protocol, established guideline values should be followed. Guideline values for concordance tend to be established for therapeutic-associated antibodies because of their complexity, eg, for which a recommended concordance of 95% should be sought. If no guidelines are published, then 90% concordance may be sufficient5 but is at the discretion of the laboratory director. The ideal validation set would include patients with known response to therapy. However, this may not be feasible within many laboratories and forms the basis of argument for specialized reference laboratories participating in relevant proficiency testing regimes (see above). The number of reference cases to be used to establish concordance depends on the degree of concordance acceptable for predictive or diagnostic purposes and the complexity of the reporting system.19 The reference standard depends on the diagnostic value of the ICC/IHC test. In the example above, in which the new antibody is part of a panel to support a diagnosis of carcinoma versus other malignancies, it would need to be compared against a collection of various cancers; a series of antibodies requires accurate concordance within a retrospective case series with cases in which status has been established.
Step 3b: What Is the Reproducibility?
This is determined by demonstrating that the same result occurs 3 times within a batch of stains and across 3 individual batches of stains over subsequent runs. It is important for both manual and automated methods of localization; eg, some manufacturers of automated systems establish reproducibility scores on installation, but these should be verified for individual antibodies if it is not cost prohibitive. Antibody lots can vary in immunoreactivity, and it should be confirmed that each lot is operating within predicted parameters against well characterized control material.
ICC—MAKING SUBJECTIVITY OBJECTIVE
The evaluation of staining quality in ICC/IHC remains subjective. In 2000, Maxwell and McCluggage,22 although they were studying a tissue section audit and quality control, recommended assessing the quality of intensity, uniformity, specificity, background, and counterstaining on an 8-point scale. Simple modification of the system allows a quick scoring methodology to assess the quality of immunoreactivity that is applicable to multiple antibodies using a common assessment system (Table 1). Kirbis et al7 argued that the use of proficiency testing or external quality-assurance systems can lead in the development of recognized standards of evaluation. Digital systems and any form of automation in pathology have focused on tissue sections.
| Staining Parameter | Comment | Value |
|---|---|---|
| Intensity | Is the immunoreactivity of adequate or strong intensity? | Negative = 0 (all other values = 0), weak = 1, moderate = 2, strong = 3 |
| Uniformity | Is the immunoreactivity distributed evenly across the material? | Uneven = 0, even = 1 |
| Specificity | Is the immunoreactivity located where expected? | Not specific = 0, specific = 1 |
| Absence of background | Is the background clear of immunoreactivity? | Excessive background = 0, background = 1, no background = 2 |
| Counterstain | Is the counterstain adequate for the intensity of immunoreactivity? | Inadequate = 0, adequate = 1 |
- a Modified from Maxwell & McCluggage 2000.22
In an analysis of ER staining on cytospin preparations, Kirbis et al23 used reference material of 2 types: cell lines that were well characterized for ER and FFPE samples that were assessed by a validated method. They also used 2 well characterized antibodies and used established methodologies for assessments. By combining well characterized materials with established methodologies, they were able to validate a methanol-fixed workflow for cytospins. More important, they demonstrated that a good understanding of preclinical parameters can provide the basis for introducing prognostic cytology for breast cancer samples.
The future of managing uncertainty is here and now. Demands from regulatory bodies in the United Kingdom and the United States for laboratories to control laboratory-development tests means that the pressure is mounting with an increase in demands for fully validated procedures.
How do we deliver demanding standards for validation? Pathologists and professional bodies have an opportunity to lead a service that is technically and scientifically aware, and they also should be experienced in morphologic and molecular testing. Training for all grades of laboratory staff in procedures and competence using equipment and in the evaluation of technical aspects of preparation is essential for establishing and maintaining such standards. Moreover, by ensuring that our trainee pathologists are equipped to combine morphology, ICC/IHC, and molecular testing, not only do we ensure the delivery of a quality service for cytology; we also provide experts in the delivery of cell-based pathology.3, 24, 25 This morphomolecular approach, which we have conceptualized elsewhere,26 forms the basis for the proposed model and, by using such a concept, a quality service for cytology can play its part.
FUNDING SUPPORT
No specific funding was utilized to prepare this commentary. In general, the Northern Ireland Molecular Pathology Laboratory is supported by Cancer Research UK, Experimental Cancer Medicine Centre Network, the NI Health and Social Care Research and Development Division, the Sean Crummey Memorial Fund, the Tom Simms Memorial Fund. and the Friends of the Cancer Centre.
CONFLICT OF INTEREST DISCLOSURES
Perry Maxwell reports support from Pfizer, Roche, and UKNEQAS during the conduct of the study. Manuel Salto-Tellez reports personal fees from PathXL, Cancer Research UK, Pfizer, Ventana Roche, and ALMAC and nonfinancial support from Visiopharm, Illumina, and Affymetrix outside the submitted work.
