Translate this page into:
Cell-blocks and immunohistochemistry
How to cite this article: Shidham VB, Layfield LJ. Cell-blocks and immunohistochemistry. CytoJournal 2021;18:2.
HTML of this article is available FREE at: https://dx.doi.org/10.25259/Cytojournal_83_2020
The interpretation of results on immunostained cell-block sections has to be compared with the cumulative published data derived predominantly from formalin-fixed paraffin-embedded (FFPE) tissue sections. Because of this, it is important to recognize that the fixation and processing protocol should not be different from the routinely processed FFPE surgical pathology tissue. Exposure to non-formalin fixatives or reagents may interfere with the diagnostic immunoreactivity pattern. The immunoprofile observed on such cell-blocks, which are not processed in a manner similar to the surgical pathology specimens, may not be representative resulting in aberrant results. The field of immunohistochemistry (IHC) is advancing continuously with the standardization of many immunomarkers. A variety of technical advances such as multiplex IHC with refined methodologies and automation is increasing its role in clinical applications. The recent addition of rabbit monoclonal antibodies has further improved sensitivity. As compared to the mouse monoclonal antibodies, the rabbit monoclonal antibodies have 10 to 100 fold higher antigen affinity. Most of the scenarios involve the evaluation of coordinate immunostaining patterns in cell-blocks with relatively scant diagnostic material without proper orientation which is usually retained in most of the surgical pathology specimens. These challenges are addressed if cell-blocks are prepared with some dedicated methodologies such as NextGen CelBloking™ (NGCB) kits. Cell-blocks prepared by NGCB kits also facilitate the easy application of the SCIP (subtractive coordinate immunoreactivity pattern) approach for proper evaluation of coordinate immunoreactivity. Various cell-block and IHC-related issues are discussed in detail.
Subtractive coordinate immunoreactivity pattern
Standard optimum cell-block protocol
SOCP NextGen CelBloking
Compared to the immunohistochemical evaluation of surgical pathology specimens, there are multiple challenges which have to be considered for the optimal outcome when interpreting the immunohistochemistry (IHC) of cell-block sections.
The most critical consideration, usually underrecognized, is potential interference associated with exposure to the various physical and chemical factors during cell-block processing. The immunoprofile of a particular neoplasm or category of cells/tissue determined by the interpretation of immunostained cell-block sections is correlated with the cumulative published database derived predominantly from formalin-fixed paraffin-embedded (FFPE) tissue sections. It is important to recognize, if the fixation and processing protocol is different from routinely processed FFPE tissue with exposure to non-formalin fixatives or reagents, the immunoprofile concluded from the immunostaining pattern of IHC on such cell-block sections may not be representative due to the potential for aberrant results.[1-3]
The continued progress in the field of IHC has been significantly advancing with the standardization of many immunomarkers with the availably of multiple novel antibodies. Introduction of various technical advances including multiplex IHCs[4,5] with refined methodologies/ automation and recent addition of rabbit monoclonal antibodies has further improved the techniques. Rabbit monoclonal antibodies demonstrate 10 to 100 fold higher antigen affinity as compared to the mouse monoclonal antibodies and have significantly higher sensitivity.
Most approaches for IHC evaluation include an interpretation of coordinate immunoreactivity patterns to decide the immunoprofile of a particular type of cell associated with a given pathology. However, as compared to tissue sections of surgical pathology specimens, the diagnostic cells and groups of cells in cell-block sections are randomly distributed. The diagnostic cells may even be difficult to categorize as the target of interest in some cases, especially in serous fluid cytology and Fine-needle aspiration biopsy (FNAB) samples with higher potential of contamination from the trajectory of the needle during endoscopic ultrasound-guided fine-needle aspiration (EUS-FNAs). Because of this, the randomly scattered cells and microtissue fragments in various random sections may be difficult to follow for coordinate immunostaining pattern in same cells in some cell-blocks.
Most of the tumor cells are large enough to be sliced into multiple 4 μm thick sections and so would be present in at least a few 4 μm thick serial levels [Figure 1]. However, such singly scattered cells may not be easily trackable in adjacent serial levels, unless the relationship of sections and the sequence of individual serial sections are properly designated on the slides and the individual sections on each glass slide are oriented exactly in an identical manner (without flipping and rotating any of the sections while transferring the paraffin sections of cell-block to the glass slides) [Figures 2 and 3].
The interpretation of coordinate immunoreactivity is based on the feature that requires evaluation and correlation of the immunostaining patterns in exactly the same cells (not just similar cells) and microtissue fragments [Figures 4 and 5]. This critical component of evaluating coordinate immunoreactivity pattern cannot be achieved in cytology preparations, because the same cell(s) cannot be present in more than one cytology preparation. This drawback, along with interferences due to the exposure of fixatives and reagents during processing of cytology preparations, does not match with those used in FFPE of surgical pathology specimen. This further discourages their application for evaluation by immunostaining as the routine primary choice. However, in some specific clinicopathological settings with specific requirements, cytology preparations may be the only option after proper standardization and optimization of protocol for a particular clinical scenario.[8-12]
Similar to the interpretation of hematoxylin and eosin (H&E) stained sections, the immunostained cell-block sections should also evaluate the immunomorphology with reference to the appropriate immunostaining patterns – membranous, cytoplasmic, nuclear, paranuclear, bile canalicular pattern etc. [Figures 5-7]. Some of these immunostaining patterns may be difficult to evaluate in cytology preparations as compared to tissue sections. Because of this, the interobserver and interinstitutional reproducibility may be compromised in the absence of methodical standardization.
The final cytopathology report should provide details as part of a standard optimum cell-block protocol (SOCP) [Figure 8] for assessing the integrity of the results on the cell-blocks prepared. Such a component in the report should assist in achieving reproducible interinstitutional results with the goal of matching with the results in published databases generated on FFPE of surgical pathology resection and biopsy specimens.
Challenges related to the evaluation of coordinate immunostaining patterns in cell-blocks are overcome by the application of subtractive coordinate immunoreactivity pattern (SCIP) approach. The remaining challenges are addressed by making cell-blocks with NextGen CelBloking™ (NGCB) kits.[1,2,7,14] Cell-blocks prepared by NGCB kits also facilitate easy application of SCIP approach, various features of which are explained below.
The relationship of each serial section of the FFPE of cell-blocks should be identifiable by sequentially numbering the glass slides with serial sections microtomed for immunostaining [Figure 1]
The serial levels of the cell-block sections are oriented identically without letting the sections be flipped, rotated, or tilted while mounting on the glass slides during microtoming by the histotechnologist [Figure 3]
Immunostaining strategies to create a basic local map of the cell-block sections [Figure 2]
Depending on the type of specimen, including Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) aspirates with bronchial and gastrointestinal (GI) mucosal contamination, and especially effusion fluids, the exact location of the diagnostic cells under scrutiny may be challenging. This may lead to suboptimal and even false-positive interpretation. For cell-blocks of effusion fluids, the SCIP approach also facilitates the creation of a basic location map of various components in the effusion fluid. SCIP approach along with vimentin immunostaining (either as single color or as dual color with BerEP4) [Figure 9] on the first serial section helps in locating all the mesothelial cells and all inflammatory cells as non-neoplastic components.
Thus, if any component which is non-immunoreactive for vimentin (and may be also BerEP4 immunoreactive in dual color immunostain) is equivalent to a second foreign population.[7,15] This approach is a very simple and effective approach for detecting metastatic disease in effusion fluid (but is usually not applicable to washings). With reference to the data in the adult population, most metastases to serous cavities are secondary to adenocarcinoma. The majority of adenocarcinomas (other than from a few sites such as renal cell carcinoma, metastatic hepatocellular carcinoma, endometrial carcinoma, etc.) are non-immunoreactive for vimentin [Table 1].[16-19]
Vimentin immunoreactive carcinomas
Renal cell carcinoma (except the chromophobe variant)
Adrenal cortical carcinoma
(Endometrial endometrioid carcinomas immunostain strongly for vimentin, but endocervical carcinomas rarely stain (weak focal staining in up to 13% of endocervical carcinomas).
Malignant mixed Müllerian tumors, serous ovarian carcinomas
Large cell carcinoma of lung
Metaplastic carcinoma of breast
Poorly differentiated adenocarcinoma of stomach (6%)
Sarcomatoid carcinomas (spindle cell carcinomas)
Pleomorphic salivary gland tumors
“Basal-like” breast carcinomas
Follicular and anaplastic thyroid carcinomas
Epithelial and sarcomatoid mesotheliomas
Solid pseudopapillary neoplasm of the pancreas
Vimentin expressing tumor epithelial cells in surgically resected carcinomas including pancreatic adenocarcinomas and hepatocellular carcinoma independently predicted a shorter post-surgical survival
Vimentin non-immunoreactive sarcoma
Alveolar soft part sarcoma
Similarly, most adenocarcinomas are immunoreactive for Ber-EP4 with a few exceptions [Table 2].[20-22] Localization and identification of these metastases are facilitated with improved efficacy by a dual color immunostaining for vimentin (as red color) with BerEP4 (as brown color), especially in cases with a small number of singly scattered metastatic tumor cells.
|Adenocarcinomas (most, 50–100% in various studies)
Neuroendocrine tumors, including small cell carcinoma
Chromophobe renal cell carcinoma (75%), papillary renal cell carcinoma (55%), clear cell renal carcinoma (18%), metastatic renal cell carcinoma (14%)
Basal cell carcinomas
Most soft-tissue sarcomas
An important caveat to this approach is that in some effusion fluids, the second foreign population with vimentin immunoreactivity may not be easily detectable. This is applicable to a relatively insignificant proportion of cases belonging to a small subset which includes a few adenocarcinomas (such as renal cell carcinoma, metastatic hepatocellular carcinoma, endometrial carcinoma, some lung carcinoma, spindle cell carcinomas from various sites, etc.) [Table 1] and most non-carcinoma malignancies (such as sarcomas, lymphomas, and melanomas). In such cases, vimentin immunoreactive tumor cells can be confirmed as non-mesothelial by evaluating their coordinate immunoreactivity in relation to mesothelial immunomarkers such as cytokeratin (CK) 7, WT1, and calretinin (nuclear).
The immunomarkers usually applied for the lineage specific evaluation include broad-spectrum anti-pan-CK antibody cocktails for carcinomas, lymphoid antigens (CD45 or CD20 and CD3) for lymphoproliferative lesions, and melanocyte differentiation antigens S100 protein (S100), Melan-A/ MART-1, tyrosinase, and SOX10) for melanoma [Tables 3 and 4]. Further evaluation of carcinomas generally includes broad categorization based on coordinate immunoreactivity for CK7 and CK20 followed by the application of lineage-specific immunomarkers [Tables 3 and 4].[16,26]
|Metastatic carcinoma (SCIP approach)|
|Broad categorization based on: Carcinoma: PanCytokeratin +
Not sarcoma: vimentin - ; Not melanoma: S100/HMB45/MelanA - ; Not hematolymphoid (lymphoma): LCA(CD45) -
Differential based on coordinate immunoreactivity pattern for CK7 and CK20
Pancreatic AdCa (subset)
Ovarian mucinous Ca
CK5/6 (-. ½+)
GCDFP, Mamaglobin, GATA-3+
TTF1, Napsin, BAP1+
mCEA, CK5/6, p63, p40-
NE markers (2 of Synaptophysin, chhromogranin,CD56), TTF1, Napsin+
Ovarian (serous) Ca
PAX8, PAX2, WT, mesothelin, ER/PR+
Ovarian (clear cell) Ca
PAX8, pVHL, HNF-1β, Napsin A+
AFP, WT1, ER, GPC3-
Maspin, S100P, MUC5AC+
CDX2, MUC1, MUC5AC−/+
Calretinin, vimentin, CK7, CK5/6, D240, WT1
BerEP4, B72.3, MOC31, mCEA
TTF1, Napsin, p63, p40, BAP1
Thyroid Ca (Non-medullary)
Thyroid Ca (medullary)
TTF1 (week to moderate)
PAX8 (variable weak)
Salivary gland (ductal) tumors
CD5+/−, p63+/−, PAX8+/−, CD117+/−, Glut1+/-
Renal (papillary) Ca
vimentin,CA-IX,RCC(Clone PN15 anti-gp200)+
Merkel cell Ca (paranuclear dot CK20)
NE immunomarkers (2 of Synaptophysin, Chhromogranin,CD56),
SATB2+, CDH17+, CDX2,TFF3,Calretinin+/−
PSA, PAP, NKX3.1,
TTF1, villin, GATA3
pCEA and CD10 with canalicular pattern
Renal (clear cell) Ca
PAX8, PAX2, vimentin, RCC,
Vimentin, inhibin, calretinin,
Yolk sac tumor
Head and Neck Ca
|Lineage category and subtypes||Immunomarkers|
|Lung adenocarcinoma||TTF-1, Napsin A|
|Breast carcinoma||GATA3, ER, GCDFP-15, Mamaglobin|
|Ovarian serous carcinoma||PAX-8, ER, WT1, vimentin|
|Ovarian clear cell carcinoma||pVHL, HNF-1B, KIM-1, PAX-8|
|Endometrial adenocarcinoma||PAX-8/PAX-2, ER, vimentin|
|Endocervical adenocarcinoma||PAX-8, p16, CEA, HPV in situhybridization, loss of PAX-2|
|Thyroid follicular cell origin||TTF-1, PAX-8, thyroglobulin|
|Thyroid medullary carcinoma||Calcitonin, TTF-1, mCEA, chromogranin|
|Salivary duct carcinoma||GATA3, AR, GCDFP-15, HER2|
|Renal cell carcinoma, clear cell type||PAX-8/PAX-2, RCCma, pVHL, KIM-1|
|Papillary renal cell carcinoma||P504S, RCCma, pVHL, PAX 8, KIM-1|
|Translocation renal cell carcinoma||TFE3|
|Upper gastrointestinal tract||CDH17, CDX2, CK20|
|Lower gastrointestinal tract||SATB2, CDX2, CK20, CDH17|
|Hepatocellular carcinoma||Arginase-1, CISH for albumin mRNA, Bile canalicular pattern (pCEA, CD10)|
|Intrahepatic cholangiocarcinoma||pVHL, CAIX|
|Pancreas, acinar cell carcinoma||Trypsin, chymotrypsin, lipase, elastase, BCL10|
|Pancreas, ductal adenocarcinoma||MUC 5AC, CK17, Maspin, S100, IMP3|
|Pancreas, neuroendocrine tumor||PR, PAX-8, PDX1, CDH17, islet-1|
|Pancreas, solid pseudopapillary tumor||Nuclear beta-catenin, vimentin, E-cadherin-Neg, PR, CD10|
|Prostate, adenocarcinoma||NKX3.1, PSA, PSAP, ERG|
|Urothelial carcinoma||GATA3, uroplakin II, S100P, CK5/6, p63, CK20|
|Adrenal cortical neoplasm||SF-1, Mart-1, inhibin-a, calretinin|
|Seminoma||SALL4, LIN28, OCT4, CD117, D2-40|
|Yolk sac tumor||SALL4, LIN28, glypican-3, AFP|
|Embryonal carcinoma||SALL4, LIN28, OCT4, NANOG, CD30, SOX2|
|Choriocarcinoma||GATA3, b-HCG, CD10|
|Sex cord stromal tumor||SF-1, inhibin-a, calretinin, FOXL2|
|Thymic origin||PAX-8, p63, CD5|
|Gastrointestinal stromal tumor||CD117, DOG1, PGDFRA|
|Solitary fibrous tumor||STAT6, CD34, Bcl2, CD99|
|Vascular tumor||ERG, CD31, CD34, Fli-1|
|Synovial sarcoma||TLE1, CK, nuclear beta-catenin|
|Chordoma||CK, S100, Brachyury|
|Desmoplastic small round cell tumor||CK, CD99, desmin, WT1 (N-terminus)|
|Alveolar soft part sarcoma||TFE3|
|Rhabdomyosarcoma||Myogenin, desmin, MyoD1|
|Smooth muscle tumor||SMA, MSA, desmin, calponin|
|Ewing sarcoma/PNET||NKX2.2, CD99, Fli-1|
|Myxoid and round cell liposarcoma||NY-ESO-1|
|Low-grade fibromyxoid sarcoma||MUC4|
|Epithelioid sarcoma||CD34, loss of INI1|
|Atypical lipomatous tumor||MDM2 (MDM2 by FISH is a more sensitive and specific test), CDK4|
|Langerhans cell Histiocytosis||CD1a, S100, Langerin (CD207)|
|Angiomyolipoma||HMB-45, SMA, Mart-1 (S100 non-immunoreactive)|
|Myoepithelial carcinoma||Cytokeratin and myoepithelial markers (may lose INI1)|
|Myeloid sarcoma||CD43, CD34, MPO|
|Follicular dendritic cell tumor||CD21, CD35|
|Merkel cell carcinoma||CK20 (paranuclear dot staining), MCPyV|
|Mesothelial origin||Calretinin (nuclear), WT1, D2-40, CK5/6, mesothelin, CK7, vimentin|
|Neuroendocrine origin||Chromogranin, synaptophysin, CD56|
|Mast cell tumor||CD117, tryptase|
|Squamous cell carcinoma||p40, CK5/6, p63, SOX2|
|Melanoma||S100, Mart-1, tyrosinase, HMB-45, MiTF, SOX10, PNL2|
iv. Orientation marker: Application of SCIP approach is facilitated further by the AV marker reported initially with Shidham’s method and included in the cell-blocks prepared by NGCB kits as a built in feature. AV marker assistance is critical while transferring various serial sections from the paraffin ribbon floated on the warm water bath. The histotechnologist during microtoming of the FFPE of cell-block can easily see the dark colored AV marker in each of the sections in the paraffin ribbon to ensure their proper orientation on sequentially numbered glass slides.
POSITIVE CONTROLS FOR IHC
As a routine approach, the sections of known positive controls are mounted on the same glass slide adjacent to the cell-block sections to be immunostained for the respective immunomarker. Different types of positive controls are described below:
Archived known tissues with positive immunoreactivity
These are the routine and most commonly used type of positive controls in most of institutions with availability of a wide range of archived tissues and neoplasms. However, the reliability of this type of positive control depends on various storage conditions including duration, storage temperature, and ambient humidity.
Multitumor tissue block (sausage method)
Multitumor tissue block prepared usually from archived FFPE with different types of neoplasms which cover a wide range of immunomarkers. Sections of such positive controls can be used for screening of hybridoma monoclonal antibodies in production settings in addition to rapid screening of novel immunomarkers in an economical manner. These, in addition to routine clinical application, can also be useful for large-scale interlaboratory quality control programs (with limitations related to the reliability applicable to the positive control mentioned above as #a).
TMA is paraffin blocks which include up to 1000 tissue cores from various types of tissues and neoplasms arranged as an array for multiplex histological analysis. Such positive controls from archived FFPE may be used to validate cancer biomarkers with large numbers of cancer samples for the evaluation of protein expression patterns by immunostaining simultaneously with identical protocols with minimal inconsistency and technical variability. TMAs can be used as IHC positive controls for prognostic, therapeutic, and diagnostic immunomarkers related to a variety of cancers (with limitations related to the reliability applicable to the positive controls mentioned above as #a and b).
Preparing FFPE of cell-blocks of formalin-fixed known standard cell lines with NGCB™ kits: Uniplex (one positive control in one section) or Multiplex (multiple immunomarkers on one section – each well in the gel disc in NextGen CelBloking™ kit with different immunomarkers (personal experience).[33a] The benefits of such positive controls are relatively consistent integrity of individual immunomarkers with better control over processing and storage conditions. This approach overcomes the limitations and problems related to the availability of a large pool of positive controls from a clinical setting in a safe and controlled manner. In comparison, the cell lines can be grown commercially as needed at production level to provide high-quality, well-standardized positive IHC controls.
IMMUNOHISTOCHEMICAL EVALUATION OF DIFFERENT TYPES OF CYTOLOGY SPECIMENS
Application of IHC to cell-blocks of Fine-needle aspiration (FNA) aspirates is relatively straight forward. It is comparable to IHC on tiny surgical pathology specimens as micro-biopsies with a few additional factors which have to be taken into consideration for the best outcome with quantitatively small amounts of diagnostic material in properly made cell-blocks.[1,2,34] Another important consideration is the potential contamination of cells/tissue material present along the track of the FNA during the procedure. This feature is significantly more frequent in procedures where the needle has to be longer such as with various EUS-FNAs. Depending on the procedure, the FNA rinses from such procedures may show variable proportions of bronchial mucosal, GI mucosal (gastric, esophageal, or duodenal), oral, and other tissue contaminants. Although these contaminations are usually easy to weed out morphologically during the interpretation, IHC interpretations on such cell-blocks could be challenging with potential pitfalls. For example, some cases with relatively few tumor cells diluted by bronchial mucosal contamination, may show many TTF-1 immunoreactive cells. This may lead to misinterpretation of the neoplasm as TTF-1 immunoreactive with potential miscategorization of tumor as adenocarcinoma. Similarly, other immunomarkers may have specific limitations in some clinicopathological situations.
A slightly different challenge is the presence of some secondary elements in the lesion with the diagnostic component of the lesion. An important example in this category is the presence of reactive lymphocytes in neuroendocrine tumors (NET) during grading of these tumors based on Ki-67 proliferation index. Some of the reactive lymphocytes aspirated during the procedure may also be immunoreactive (nuclear) for Ki-67 and may be counted as NET cells and lead to a higher count with false upgrading of NET, especially for tumors in borderline categories. This challenge can be overcome by applying dual color IHC (Ki-67: Brown, nuclear with LCA: Red, cytoplasmic).
Immunohistochemical evaluation of hematolymphoid pathologies
Similarly for lymphoproliferative neoplasms and pathologies, the interpretation is dependent on the immunoprofile of various components in the lesion. In general, flowcytometry is the preferred approach for the evaluation of immunoprofiles in these pathologies. However, availability of cell-blocks in these situations is crucial for the evaluation of some immunomarkers which are better evaluated by IHC. In addition, cell-block may be required to perform some molecular pathology tests (for more details refer to other review articles on this topic).
Immunohistochemical evaluation of effusion fluids
As compared to IHC on other type of specimens, application and interpretation of IHC on cell-block sections of effusion fluids is relatively complex due to a wide morphological spectrum of reactive mesothelial cells.[7,35-37] Although the cytomorphological clues helping to identify the second foreign population of tumor cells may be facilitated by DiffQuik stain, interpretation and confirmation are relatively objective with IHC. The SCIP approach with vimentin (one color or dual color) immunostaining with BerEP4 detects the second foreign population of tumor cells in most of the cases with significant ease and reproducibility.
However, the interpretation of immunostained cell-block sections of effusion specimens with relative hypocellularity of diagnostic cells diluted with reactive mesothelial and inflammatory cells is challenging. This is further accentuated if the metastasis is from a well to moderately differentiated carcinoma presenting as scattered isolated tumor cells such as in metastatic mammary carcinoma (especially lobular carcinoma), ovarian carcinoma, and some lung carcinomas. The same challenge may also compromise the selective microdissection of the tumor cells for molecular pathology studies.
Conventionally, vimentin is dismissed as a non-specific immunomarker. However, it has a very critical role in identification of a second foreign population of metastatic tumor cells in serous fluid cytology in the adult population. This benefit of vimentin immunostaining in concert with the SCIP approach may be applied as single color or dual color technique (vimentin red with BerEP4 brown) for objective detection of a second foreign population or tumor cells in cell-block sections of effusion fluid. In most of the cases, they would be detected as a vimentin non-immunoreactive population with Ber-EP4 immunoreactivity [Figure 9].[38,39]
Depending on the cytomorphology and clinical setting, the differential interpretations for effusion fluid IHC should also include mesothelioma, primary peritoneal carcinoma, and primary effusion lymphoma.[40,41] The approach of identifying the second foreign cell population will not assist in effusions from cases with the epithelioid variant of mesothelioma because of the overlap of immunoprofile with that of reactive mesothelial cells. In such cases, additional quantitative (numerous vs. a few) and qualitative (numerous large groups vs. a few small groups) features would help along with clinical and radiologic findings of mesothelioma with approach to rule out metastatic adenocarcinoma and application of some recently available molecular pathology tests for mesothelioma.
In contrast to most of the adenocarcinomas (except a few tumors such as renal cell carcinoma, some lung adenocarcinoma, Mullerian tumors, and metastatic hepatocellular carcinoma), most of the mesothelioma cells and reactive mesothelial cells are immunoreactive for vimentin [Table 1]. The mesothelial cells are also immunoreactive for CK 7 and calretinin (nuclear and cytoplasmic). However, the immunostaining pattern of mesothelioma for vimentin immunostaining may be heterogeneous with focal vimentin non-immunoreactivity. Other immunomarkers for mesothelial cells include WT1, D2-40, and CK 5/6 in positive category. In addition to confirming the cells as mesothelial cells, depending on a particular clinical setting, potential metastasis from other primary sites in the differential diagnosis should be ruled out with the immunomarkers in the negative category for mesothelial cells (including some relatively organ-specific immunomarkers such as CDX2 (colorectal and upper GI), TTF-1 (lung), Napsin 1 (lung), PAX8 (Mullerian and TTC), GATA3 (breast), ER (breast and Mullerian), PSA (prostate), PAP (prostate), AFP (germ cell tumors), TFE (alveolar soft part sarcoma), HCG (choriocarcinoma), S100 (melanoma and MPNST), and synovial sarcoma (beta-catenin and SYTSSX) [Tables 3 and 4].
IHC ON CELL-BLOCKS TREATED WITH FORMIC ACID FOR PRIONS
The cell-block sections of formic acid treated FFPE will require pretreatment or combination of pretreatments for IHC analysis for the best results. Based on the study of IHC for the prion protein on brain tissue of CJD patients, hydrated steam autoclaving (HA) for 10 min at 121°C in 10 mM citric acid recovery buffer at pH 6 showed best results. However, for other immunomarkers, it is recommended to standardize this antigen retrieval protocol with a positive control processed in the same manner with formic acid treatment similar to the formic acid treated FFPE of the cell-block.
DIFFERENTIAL DIAGNOSIS OF UNKNOWN PRIMARY
Unknown primary may be defined as a clinicopathologic scenario when the exact site of origin of a neoplasm cannot be determined based on routine evaluation with clinical and imaging studies.[16,23,39,41,45-50] Knowing the exact nature of the primary site of origin is critical for proper treatment and management of the malignancy. In brief, it is an occult primary with metastatic spread and needs a tissue diagnosis with the help of ancillary studies such as IHC and molecular pathology testing. There are numerous publications which discuss pathological evaluation of unknown primary.
In general, the nature of the primary site is usually obvious in cases with metastases to serous cavities as Stage IV phenomenon. Because of this, in majority of cases, cytopathologic and IHC evaluation may usually require confirmation of a second foreign population as a metastasis. However, in some cases, IHC may have to be extended to confirm the primary site of an unknown primary or for differential diagnosis between two or more possible primaries in some rare cases. The challenge may be complex in such cases with an approach to locate the metastatic tumor cells and track them by additional immunomarkers (under SCIP approach) depending on possible differential diagnosis for the evaluation of the primary site [Tables 3 and 4].
NGCB™ kits which has a built-in AV marker result in cell-blocks which allow improved interpretation with the SCIP approach [Figure 2]. Combination of various immunomarkers including site-specific immunomarkers [Tables 3 and 4] which allow interpretation of coordinate immunoreactivity for two and non-immunoreactivity for two tissue-specific immunomarkers for the same tumor cells would result in higher accuracy.
Most of the other specimens, including various FNAs, require methodical application of IHC for the evaluation of an unknown primary. Additional methods, including molecular techniques, may be needed as indicated. The approach begins with simple initial steps to broadly categorize the malignancy as carcinoma, sarcoma, hematolymphoid process – lymphoma, or melanoma [Tables 3 and 4].[25,48] This step is relatively straightforward in most cases with IHC.
Continued progress in IHC has significantly advanced with availability of many organ- and tumor-specific immunomarkers including multiplex IHCs.[4,5] Determination of primary site without any prognostic or therapeutic advantages would equate with expensive and unnecessary testing. However, diagnosing the primary site is critical for delivering specific treatment for an increasing number of primary sites. Due to this, precision in determination of primary site is very critical in the current era with ongoing progress in molecular and immuno-based targeted therapies.
Site-specific treatment options are relatively more effective for some metastases such as from breast, ovary, prostate, and thyroid with improved prognosis with ongoing additions of other sites including colorectal and lung. IHC in correlation with clinical and imaging findings usually identifies the primary in more than 90% of the cases. In the remaining cases, application of molecular analysis may help to identify the primary sites.
There are many novel site-specific immunomarkers with ongoing additions for interpretation of a primary site [Tables 3 and 4]. Many publications review and suggest various algorithms and guidelines.[16,38,39,41,43,49,50,54] Many more immunomarkers will be added in future with an ever increasing number of site-specific and therapy-specific immunomarkers [Table 5]. However, in some cases, the results may not be conclusive because of scant or poorly preserved or improperly processed specimen. In such cases, performing IHC on repeat specimen with adequate proportion of diagnostic tumor cells may be conclusive in correlation with clinical and imaging findings. In some unresolved cases, other ancillary tests such as molecular studies may be required.[45,48,55-60] (see chapter #5).
SOME IHC-RELATED PITFALLS
Non-specific immunostaining due to a proteinaceous background around the diagnostic components in the cell-block sections may interfere with proper evaluation of the immunoreactivity pattern in immunostained cell-block sections.
Distribution pattern of diagnostic components in the cell-block sections (tiny cell groups vs. single cells), fixative(s) used, choice of antigen retrieval protocol (i.e., heat-induced epitope retrieval, enzyme digestion, etc.), titer of antibody used, clone of the antibody, and other similar potential variables in the protocols may interfere with the immunostaining pattern.
Immunoexpression of CKs, although usually associated with epithelial neoplasms, is known to occur in some noncarcinomas. The list of such tumors includes synovial sarcoma, adamantinoma, chordoma, epithelioid sarcoma, myoepithelial neoplasms, and desmoplastic small round cell tumor. Similarly, focal aberrant immunoexpression for CK may be observed in some tumors which are not expected to demonstrate CK immunoexpression. There are a relatively significant number of tumors in the latter category. To mention, a few of relative significance are melanomas, peripheral nerve sheath tumors, angiosarcomas, meningiomas, clear cell sarcomas, undifferentiated pleomorphic sarcomas, epithelioid hemangioendotheliomas, epithelioid leiomyosarcomas, plasmacytomas, rhabdomyosarcomas, malignant rhabdoid tumors, and even lymphomas such as diffuse large B-cell lymphomas and anaplastic large cell lymphomas.
As compared to the primary tumor, the site-specific immunomarkers in metastases may show a complete loss or relatively weak immunoreactivity for the immunomarker. This feature of commonly observed heterogeneous immunoexpression pattern in some tumors may cause sampling of non-immunoreactive areas leading to misinterpretation due to sampling artifact. This challenge has a relatively higher significance in cell-block sections with relatively a few diagnostic cells and if interpretation is based on a single or few immunomarkers. Thus, it is important to practice methodologies achieving best cellularity and qualitative integrity of cell-blocks possible with NGCB™ kits along with the practice of applying multiple immunomarkers sufficient to evaluate at least two positive and two negative immunomarkers for a particular tumor in the differential interpretation.
Relatively underestimated but an equally significant issue is improper processing related aberrations leading to the potential compromisation of patient care with related liability. The study evaluating effects of fixation and the time in relation to IHC on cell-block sections summarized that formalin preservation should be the preference for proper results of IHC. Most importantly, alcohol preservation/ fixation introduces an unpredictability of immunoreactivity usually with reduced immunostaining depending on how long the cells/tissue is exposed to alcohol. The pre-analytic variables are critical for proper IHC results with an important role of proper positive controls. A study evaluating PD-L1 immunostaining of cell-block sections with reference to various processing approaches, reported that formalin, RPMI, and saline demonstrated stronger cytoplasmic immunostaining as compared to Cellient cell-blocks, with the poorest immunostaining with CytoLyt.[63-66]
Formaldehyde has preferential interaction with primary amines (such as lysine in proteins) and primary amides (asparagine and glutamine) present in tissue for intra- and intermolecular covalent cross-linking of various biomolecules, which may impair the enzymatic activity and also immunoreactivity.[67,68] Due to this, it requires proper application of the antigen retrieval step on FFPE sections of cell-blocks similar to FFPE sections of surgical pathology specimens.
The immunoprofile of the tumor cells or diagnostic components in cell-blocks prepared with methodologies which do not match with FFPE processing would be interpreted by comparing with immunoprofiles in published data performed on FFPEs of surgical pathology biopsy specimens. Because of this, the results generated would be non-representative and aberrant with ultimate compromisation of delivery of appropriate therapy with a direct impact on patient care and related liability for the testing institution. However, it is not uncommon to apply protocols in which the specimens are collected in various alcohol-based fixatives such CytoLyt™, PreservCyt® (ThinPrep), or CytoRich™ Red preservative (SurePath).[2,63,70,71] As mentioned previously, this introduces significant liability including malpractice and other legal concerns in some settings such as mesothelioma claims. This pitfall can be easily prevented by applying the protocol similar to that used for surgical pathology specimens which can be achieved with the help of recently available NGCB™ kits.[1,33]
IHC pitfalls and challenges with reference to effusion fluid cell-blocks
As compared to the interpretation of immunostained cell-block sections of FNA specimens, evaluation of immunostained cell-block sections of effusion fluids poses a significant number of special challenges. Some unexpected aberrant immunoreactivity may lead to incorrect interpretations.
|Estrogen receptor (ER)||The ER and PR status should be tested on the primary tumor and/or metastases for all newly diagnosed invasive breast cancer or recurrence
For ductal carcinoma in situ, ER status is required to decide hormone therapy to reduce the risk of future breast cancer. Testing DCIS for PR status is optional
|Progesterone receptor (PR)|
|HER2/neu (ERBB2; c-erbB-2; erb-b2 receptor tyrosine kinase 2; human epidermal growth factor receptor 2)||For application of HER2-targeted therapy
Tested in invasive or recurrent breast cancers, both lobular and ductal (Joint ASCO and CAP)
Similarly, inoperable, locally advanced, recurrent or metastatic stomach and esophagus cancers (Joint ASCO and CAP)
HER2 status tested by: (a) Immunohistochemistry to measures HER2 protein. (b) Fluorescent in situhybridization evaluates amplification (increased number of gene copies)
|CD117||Majority of GISTs overexpress CD117 (c-KIT) and PDGFRA. DOG1 (anoctamin-1/ANO1, a voltage-gated calcium-activated anion channel) immunoexpression may be noted even in CD117 non-immunoreactive GISTs. A minority of GISTs only shows PDGFRA immunoreactivity SDHB immunohistochemical staining should be considered for any tumors that lack CD117 or PDGFRA immunoreactivity to identify SDH-deficient GIST|
|EGFR||EGFR is highly expressed in a variety of solid malignant tumors (including non-small cell lung cancer, pancreatic cancer, breast cancer, medullary thyroid cancer, salivary gland carcinoma, squamous cell carcinoma of the head and neck, colorectal cancer, chordoma, and malignant gliomas) and its expression has been correlated with disease progression and poor survival
With increasing role of targeted therapies such as Anti-EGFR monoclonal antibody (IMC-C225: Cetuximab) IHC-based EGFR screening methods using FFPE of tumor assists selection of cancer patients eligible for cetuximab treatment
|p16||p16 immunohistochemistry for risk stratification in oropharyngeal SCC with significantly better outcome of p16-positive oropharyngeal SCC than for p16 negative tumors|
|p40||Thyroid transcription factor-1 (TTF1) and p40 (an isoform of p63) are immunomarkers for adenocarcinoma (ADC) and squamous cell carcinoma (SCC) respectively for objective categorization (especially in cases with solid/squamoid morphology which may be misclassified as SCC with therapy (with bevacizumab) related potentially lethal pulmonary hemorrhage|
|p53||p53 is altered gene in human cancers in approximately 50% of all invasive tumors. The most difficult-to-treat cancers (such as high-grade serous ovarian cancers, triple-negative breast cancers, esophageal cancers, small-cell lung cancers and squamous cell lung cancers) show p53 mutation in at least 80% of cases
In such cases, mutant p53 protein is a candidate target for the new anticancer therapies
|Ki67||For grading of neuroendocrine tumors, GIST, pheochromocytoma, lymphoma
As proliferation marker could be important in many other tumors including breast carcinoma
|Ber-EP4||Cases with metastasis of Ber-EP4 immunoreactive (epithelial cell adhesion molecule -positive tumors) may be treated with intraperitoneal catumaxomab antibody|
|MMR proteins (MLH1, MSH2, MSH6, and PMS2)||To identify patients at risk for Lynch syndrome(hereditary non-polyposis colorectal cancer) and patients with sporadic microsatellite instable colorectal cancer
Lynch syndrome is associated with increased risk of carcinoma at various sites including:
Gastrointestinal: Colon, small intestine, stomach, hepatobiliary, pancreas
Genitourinary: Kidney, bladder, prostate
Gynecological: Endometrium, ovary
|PDL1||Therapeutic monoclonal antibodies that target either PD-1 or PD-L1 have been FDA approved for use in various malignancies (including metastatic melanoma, non-small cell lung cancer (NSCLC), renal cell carcinoma, bladder cancer, head and neck cancer, Merkel cell carcinoma, Hodgkin lymphoma, gastric cancer, hepatocellular carcinoma, and microsatellite instability-high cancer regardless of histology). Approval is pending in other diseases
However, testing for PD-L1 positivity is required for therapy of NSCLC with pembrolizumab (PD-1/PD-L1 antibody). However, there are four FDA registered PD-L1 IHC assays based on four different PD-L1 antibodies (22C3, 28-8, SP263, and SP142) used on two different IHC platforms (Ventana and Dako) with different scoring systems
BerEP4 immunostaining in some mesothelioma cells may lead to their interpretation as metastatic adenocarcinoma. BerEP4 immunostaining in mesothelioma cells if present is usually membranous with microvillous pattern. This could be evaluated more clearly under an oil immersion lens at higher magnification [Figures 7 and 10]. Immunomorphological evaluation, instead of just observing positive or negative immunostaining, is critical in such cases. Contrarily, a few metastases may be Ber-EP4 non-immunoreactive [Table 2] but detectable as a foreign second population due to vimentin non-immunoreactivity. Such cells in effusion fluids are consistent with metastatic tumor cells, which should be evaluated properly by following their coordinated immunostaining pattern for other relevant immunomarkers.
Similarly, focal, faint immunoreactivity for Napsin A in some reactive mesothelial cells may be observed with dual color immunostaining without immunoreactivity for nuclear TTF1 in rare cases (personal experience). However, this could be avoided with the inclusion of an adequate number of immunomarkers in the panel with a coordinate immunostaining pattern as two negatives and two positives in the same cells evaluated by the SCIP approach.
As observed in some cases, without a known malignancy or metastasis (personal experience), the predominance of small chronic inflammatory cells with a scant amount of cytoplasm may show negligible vimentin immunoreactivity which may be misinterpreted as a vimentin non-immunoreactive population with potential for false-positive interpretation as metastasis. Vimentin non-immunoreactivity in some cells interpreted as a second foreign population can be avoided by a careful morphological evaluation at higher magnification. The small cells can be confirmed as lymphocytes with minimum amount of cytoplasm with scant focal vimentin immunoreactivity in at least a few cells which can be distinguished from the larger tumor cells with relatively more cytoplasm [Figure 11].
Scant cellularity may compromise the ability to interpret and follow the same cell (or small groups of cells) for the evaluation of coordinate immunoreactivity, primarily due to a lack of identical orientation between the serial sections. Routine application of the SCIP approach would overcome this pitfall in a significant proportion of cases.
Predominance of singly scattered tumors cells simulating a single population (especially in cases with tumors cells showing cytomorphology overlapping with reactive mesothelial cells) may lead to a false-negative interpretation of H&E stained cell-block sections of effusion fluids. This is significantly more challenging with metastases from lower grade tumors. The list of such primary sites includes the breast (lobular carcinoma) and low-grade ovarian neoplasms. These scenarios can be avoided by the SCIP approach with IHC of the first section for vimentin as one color or as dual color (vimentin: Red, cytoplasmic with BerEP4: Brown, cytoplasmic) immunostaining. This approach would easily achieve objective confirmation in most of cases, even those with relatively scant numbers of diagnostic tumor cells. With reference to the dual color immunostaining (vimentin: Red, cytoplasmic with BerEP4: Brown, cytoplasmic), in some rare cases, the tumor cells may show immunoreactivity for both vimentin and BerEP4 which may be difficult to interpret. In such cases, IHC may be repeated as single color immunostaining for vimentin and BerEP4, respectively.
CellBlockistry, the science and art of art of cell-block making, refines cell-blocking of variety of cytology specimens. However, currently, multiple methods have been practiced with significant variation with lack of reproducibility in qualitative and quantitative parameters of final cell-blocks. For the best outcome, the fixation and processing of cell-block should be comparable to FFPE of biopsy and surgical pathology specimens. Such a practice in universal fashion would result in reproducible results with interinstitutional comparability. FFPE of such cell-blocks would yield results of ancillary studies (especially IHC and molecular tests) comparable to those of FFPE of biopsy and surgical pathology specimens. This is highly significant due to the increasing number of tests with direct impact on selection of targeted therapy in the era of personalized medicine.
|Every cytology report on cytology specimens with cell-block should have following minimum details communicated in it under gross description section or other designated section such as quality details
This would allow proper decision-making in relation to various quality related aspects when any ancillary tests are performed
Number of cell-blocks prepared with their designation (with any descriptive comments similar that in surgical pathology report): ______
A1 (prepared from the clot in fresh unfixed specimen)
A2 (from sediments after lysis of red blood cells with
(mention method with reference if possible), etc.
Specimen collected in
Saline/RPMI/Hanks balance solution/Isotonic Medium S™/other ______
Other (not recommended due to potential interference with the results of variety of IHC and molecular test results): CytoLyt/Saccamanno’s fixative/CytoRich Red/other alcohol based or acid based non-formalin reagents
Duration of specimen in the collection medium and temperature
Prior to actual fixation in 10% formalin
Temperature: 2–8°C/room temperature /Other _______________
Any processing before making the cell-block (and before final fixation in 10% formalin)
Lysis of red blood cell contamination: With (mention method used)
CytoRich Red®(potential interference with IHC), BloodLyz™, other lysing reagent _______
Fixation time in 10% formalin (before start of actual tissue processing)
As a routine, cell-block is recommended to be made from the residual cytology specimens after making cytology preparations. Cell-blocks are critical for performing various elective ancillary studies including IHC, special stains for organisms or chemical components, and the molecular studies. Some of these elective studies may not be required immediately during the time of cytopathology reporting, but such tests may be indicated at later stage of the disease. In such situations, cell-block will be an excellent archival material which could be used any time at a later date.
The entire process starts with a proper specimen collection. The preferred approach is to collect it as fresh and unfixed specimen which permits the flexibility of using the recommended protocol matching that used for biopsy/ surgical pathology specimens [Figure 12]. Shidham’s method addresses many of the quantitative concerns including guidance to histotechnologists for the selection of sections with diagnostic material with the help of AV marker. However, this method as in its original published protocol is heavily dependent on skill and labor. Due to this, it is difficult to be adopted as a routine by individual laboratories. Recently, a ready-to-use, low cost, alternative based on Shidham’s method is available as NGCB™ kits which overcome the limitations with initial protocol. The cell-block making units in the kits include a built-in, precisely set, dark colored AV Marker to monitor the depth of cutting objectively by the histotechnologists microtoming the FFPE of cell-blocks.
In addition to standardization for making quantitatively and qualitatively optimum cell-blocks, the final cytopathology report should document the critical SOCP information in the form of quality indicator component in the final cytology report [Table 6]. The introduction of this practice is critical to increase the impact of cell-blocks in patient care with enhanced reproducibility and comparability of results of important ancillary tests such as IHC performed on the cell-blocks.
Authors thank Drs. Ahmed Alrajjal and Mir Yousuf Khan for contributing images for some figures. Authors also thank Janavi Kolpekwar, Kathy Rost, and Dr. Moumita Saha Roy Choudhury for their copy editing support.
COMPETING INTERESTS STATEMENT BY ALL AUTHORS
VS (the first author) and LL (the second author) are the coeditors-in-chief of CytoJournal. They do not have any competing interests. The spouse of VS has stakes in AV BioInnovation LLC product(s) cited in this review.
LIST OF ABBREVIATIONS (In alphabetic order)
CK – Cytokeratin
CMAS – CytoJournal Monograph & Atlas Series
EUS-FNA – Endoscopic ultrasound-guided fine-needle aspiration
FFPE – Formalin-fixed paraffin embedded
FNA – Fine-needle aspiration
FNAB – Fine-needle aspiration biopsy
HA – Hydrated steam autoclaving
GI – Gastrointestinal
H&E – Hematoxylin and eosin
IHC – Immunohistochemistry
NET – Neuroendocrine tumor
NGCB – NextGen CelBloking™
S100 – S100 protein
S100P – Placental S100
SCIP – Subtractive coordinate immunoreactivity pattern
SOCP – Standard optimum cell-block protocol
TMA – Tissue microarray
- CellBlockistry: Science of cell-block making as ancillary cytopathology component in the era of minimally invasive techniques with increasing role of molecular pathology (short communication) Clin Surg. 2019;4:2510. Available from: http://www.clinicsinsurgery.com/pdfs_folder/cis-v4-id2510.pdf [Last accessed on 2021 Jan 06]
- [Google Scholar]
- CellBlockistry-IAC 2019 (May 6-Monday 16-1500 to-1800) Chemistry and Art of Cell-Block Making (Part I) 2019. Available from: https://www.slideshare.net/vshidham1/01-cellblockistry-iac-2019-may-6monday-16-1500-to-1800 [Last accessed on 2021 Jan 06]
- [Google Scholar]
- Immunocytochemistry of effusion fluids: Introduction to the SCIP approach In: Shidham VB, Atkinson BF, eds. Cytopathologic Diagnosis of Serous Fluids (1st ed, Ch. 5). Amsterdam: Elsevier, WB Saunders Company; 2007. p. :55-78.
- [Google Scholar]
- CellBlockistry-IAC 2019 (May 6-Monday 16-1500 to-1800) Chemistry and Art of Cell-Block Making (Part II) 2019. Available from: https://www.slideshare.net/vshidham1/02-cellblockistry-iac-2019-may-6monday-16-1500-to-1800 [Last accessed on 2021 Jan 06]
- [Google Scholar]
- Efficiency and Effectivity of Using Dual-Color Immunostaining with Combination of BerEP4 and Vimentin as Compared to BerEP4 Alone on Serous Effusion, 68th Virtual Annual Scientific Meeting. American Society of Cytopathology.
- [Google Scholar]
- Undifferentiated tumor: True identity by immunohistochemistry. Arch Pathol Lab Med. 2008;132:326-48.
- [Google Scholar]
- BerEP4/EpCAM. 2020. Available from: https://www.pathologyoutlines.com/topic/stainsepcam.html [Last accessed on 2020 Nov 27]
- [Google Scholar]
- Hematolymphoid disorders In: Shidham VB, Atkinson BF, eds. Cytopathologic Diagnosis of Serous Fluids (1st ed, Ch. 12). Amsterdam: Elsevier, WB Saunders Company; 2007. p. :171-93.
- [Google Scholar]
- Positive Control Tissue List for Immunohistochemistry. 2020. Available from: http://www.ihcworld.com/_technical_tips/positive_control_tissue_list.htm https://www.archive.is/ksw4T [Last accessed on 2020 Apr 18]
- [Google Scholar]
- The multitumor (sausage) tissue block: Novel method for immunohistochemical antibody testing. Lab Invest. 1986;55:244-8.
- [Google Scholar]
- Immunohistochemistry: Diagnostic and prognostic applications In: Detrick B, Hamilton RG, Folds JD, eds. Manual of Molecular and Clinical Laboratory Immunology (7th ed, Ch. 47). Washington, DC: American Society of Microbiology Press; 2006. p. :408-13. Available from: https://www.tetondata.com/TitleInfo.cshtml?id=274 [Last accessed on 2021 Jan 06]
- [Google Scholar]
- Pleural, peritoneal, and pericardial fluids In: Bibbo M, ed. Comprehensive Cytopathology (2nd ed). Philadelphia, PA: WB Saunders; 1997. p. :551-621.
- [Google Scholar]
- Shandon™ CytoRich™ Red Collection Fluid. Available from: https://www.thermofisher.com/order/catalog/product/b9990802#/b9990802 [Last accessed on 2021 Jan 06]
- [Google Scholar]
- Mesothelioma In: Shidham VB, Atkinson BF, eds. Cytopathologic Diagnosis of Serous Fluids (1st ed, Ch. 8). Amsterdam: Elsevier, WB Saunders Company; 2007. p. :107-13.
- [Google Scholar]
- Cancer of unknown primary site: Evolving understanding and management of patients. Clin Adv Hematol Oncol. 2012;10:518-24.
- [Google Scholar]
- Genomic Testing, FoundationOne®. Cambridge, MA, USA: Foundation Medicine; Available from: https://www.foundationmedicine.com/genomic-testing [Last accessed on 2021 Jan 06]
- [Google Scholar]
- Immunohistochemistry on Cell-block Specimens Effects of Fixation and Time (PowerPoint Presentation of a Study) 2014. British Columbia, Canada: BC Cancer Agency Clinical Laboratory Vancouver; Available from: http://www.cpqa.ca/main/wp-content/uploads/2014/06/2014-thomson.pdf [Last accessed on 2020 Apr 20]
- [Google Scholar]
- ThinPrep Non-Gyn. Available from: http://www.thinprep.com/hcp/lab_professionals/thinprep_non_gyn.html [Last accessed on 2021 Jan 06]
- [Google Scholar]