Precision Medicine in Cytopathology

Abstract

Over the last decade, cancer diagnostics has undergone a notable transformation with increasing complexity. Minimally invasive diagnostic tests, driven by advanced imaging and early detection protocols, is redefining patient care and reducing the need for more invasive procedures. Modern cytopathologists, beyond their diagnostic role, now safeguard patient samples for vital biomarker and molecular testing. In this article, we explore ancillary testing modalities and the role of biomarkers in organ-specific contexts, underscoring the transformative impact of precision medicine. Finally, the advent of more than 80 FDA-approved predictive biomarkers signals a new era, guiding cancer care toward personalized and targeted strategies.

Overview:

Over the past decade, the complexity of cancer diagnostics has rapidly expanded. The number of newly defined tumor types and subtypes that have unique morphology, immunochemical protein expression, and molecular findings has dramatically increased and has led to better risk stratification and treatment options for patients in most primary sites.¹ Simultaneously, there have been a number of advancements regarding protein expression and molecular drivers that have been found to predict response to targeted therapies.^2–4 Despite this increased complexity, we have been progressively able to do more with less tumor sample.^5–7 With the advent of high-resolution CT, multi-sequence MRI, and early detection cancer screening protocols, patients often rely on minimally invasive diagnostic tests to target previously un-targetable lesions, obviating a need for a more invasive or morbid procedure.^{8, 9} With these samples, we have moved beyond the need to merely provide an accurate diagnosis. The modern cytopathologist must be a good steward of the patient’s sample and preserve the specimen for relevant biomarkers and molecular testing when indicated. This role requires the cytopathologist to have a functional understanding of the specimen needs for each accompanying diagnosis and clinical scenario. Fortunately, many of these studies can be performed with relatively low tumor fractions and low minimum nucleic acid quantities.¹⁰

When discussing biomarkers, a definition of terms is warranted. First, diagnostic markers are used for the purpose of rendering a primary diagnosis. These include immunostains such as p40 or TTF-1 that help support squamous differentiation or a lung or thyroid primary site of disease, FISH studies that detect MDM2 amplification in a well-differentiated liposarcoma, or RNA sequencing that detects a NUTM1 fusion in a NUT carcinoma, to name a few.^11–14 Next, prognostic markers provide information about the likely course of a disease, regardless of treatment. Essentially, these markers help predict the natural history of the disease or the overall outcome. These include markers such as Ki-67, ATRX, and DAXX in pancreatic neuroendocrine tumors, POLE-e and p53 in endometrioid carcinomas, and microsatellite instability in colon cancer, to name a few.^15–18 Finally, predictive markers essentially forecast the likelihood of benefit for a particular therapy. Examples include ER/PR and HER2 expression in breast cancer, ALK gene fusions in lung adenocarcinoma, and PD-L1 expression and tumor mutational burden in many solid tumor types.^{4, 19–22} These predictive biomarkers can be further subdivided into tumor specific markers, meaning the correlation of therapy depends on the specific tumor type, and tumor agnostic markers, meaning response to therapy is independent of the specific diagnosis. For example, an EGFR mutation in lung adenocarcinoma would be tumor specific while an NTRK gene fusion in any solid tumor would be tumor agnostic and considered actionable in the proper clinical setting.^{23, 24} Due to the interdisciplinary complexity in integrating all of this information, some centers are developing molecular tumor boards to improve clinical care.^{10, 25}

As laboratory physicians, it is crucial for the cytopathologist to have a solid understanding of these ancillary and molecular tests so we can adequately and accurately answer clinical care questions for our clinicians. This understanding further affects how we handle rapid onsite evaluation cases, triage specimens, and make procedural recommendations to clinicians. In this review, we will provide an overview of the current major ancillary testing modalities, the role of molecular diagnostic testing for clinical decision-making in cytopathology, and precision medicine biomarkers that can be performed in exfoliative, aspiration, and small core needle biopsy samples.

Discussion:

Ancillary Testing Modalities

There a several important methodologies that are utilized for detecting biomarkers in cytopathology samples. These include immunochemistry, fluorescence in situ hybridization (FISH), chromogenic in situ hybridization (CISH), microsatellite instability, DNA sequencing, and RNA sequencing, in addition to polymerase chain reaction (PCR testing) for human papillomavirus testing in gynecologic cytology samples. Test selection depends on the specimen type, the specific biomarker, and the specimen requirements.

Immunochemistry

Briefly, immunochemistry relies on the use of antibodies to detect the presence, abundance, and localization of specific proteins in the sample and can be performed on both alcohol-fixed smears, liquid-based preparations, and cytospins, as well as more commonly on formalin-fixed paraffin embedded (FFPE) cell blocks and core needle biopsies. However, several of the predictive biomarkers have specific preanalytical variable requirements that are important to be aware of. For example, breast biomarkers require 10% neutral buffered formalin fixation for 6–72 hours because both under- and over-fixation can result in false negatives.^{26, 27} In addition, some drugs require a specific clone of antibody for predictive purposes (Figure 1A–D). For example, the PDL1 clone 22C3 is an FDA approved companion diagnostic test for pembrolizumab in certain clinical settings for patients with non-small cell lung carcinoma, cervical squamous cell carcinoma (SqCC), endocervical adenocarcinoma, urothelial carcinoma, head and neck and esophageal SqCC, and triple negative breast cancer; however, the scoring method and the threshold for positivity is tumor specific (Figure 1E–F).^28–33 In contrast, the PDL1 SP142 clone is an FDA approved companion diagnostic test for atezolizumab only in patients with urothelial carcinoma and advanced non-small cell lung carcinoma.^{34, 35} Due to these differences, each clone and each primary site has to be separately validated. As a result, it is important to be aware of which clone or clones your lab has inhouse and to be aware of which specific primary tumor types have been validated. Wasting tissue on an improperly ordered PDL1 for which your lab is not validated may deplete the tumor cellularity and require the patient to undergo a repeat procedure. In other settings, immunochemistry can be used as a quick triage technique. For example, immunochemistry for ALK and ROS1 in non-small cell lung cancer can be performed to identify cases that can subsequently be confirmed to have a gene rearrangement by FISH, PCR, or sequencing studies, making the patient eligible for a tyrosine kinase inhibitor.

Figure 1.

A. Metastatic ductal carcinoma of breast, H&E 20X; cell block section from station 7 lymph node endobronchial ultrasound-guided fine needle aspiration biopsy. B. Metastatic ductal carcinoma of breast, ER immunostain 20X, interpreted as moderate intensity staining in 61–70% of tumor cells. C. Metastatic ductal carcinoma of breast, PR immunostain 20X, interpreted as moderate intensity staining in 1–10% of tumor cells. D. Metastatic ductal carcinoma of breast, HER-2/neu immunostain 20X, interpreted as positive (score 3+). E. Adenocarcinoma of lung, H&E 20X, CT-guided core needle biopsy from right upper lobe of lung. Adenocarcinoma of lung, PD-L1 IHC 22C3 pharmDx 20X, interpreted as 60% tumor proportion score.

Fluorescence in situ hybridization (FISH)

FISH studies use fluorescent probes to visualize the present and location of a specific DNA sequence and can be performed on cytologic slide preparations as well as FFPE cell blocks and core biopsies for diagnostic and predictive purposes. As examples, FISH can be used to detect HER2, c-MYC, and MDM2 gene amplifications, deletions such as 1p/19q codeletions, and fusions such as in ALK, ROS1, RET, SS18-SSX, and many more.^{13, 20, 23, 36–40} There are several different ways to perform a FISH assay depending on the type of aberration that you need to detect. The most frequently utilized types of assays for solid tumors include break-apart, dual fusion, and centromere enumeration probe FISH. Briefly, break-apart FISH detects chromosomal rearrangements by labeling and hybridizing to two separate regions of a target gene. The test is considered to be positive for a fusion if the distance between the probe signals is far apart, correlating with a break and rearrangement of a gene, irrespective of the fusion partner (Figure 2). Dual fusion FISH on the other hand utilizes two differently colored probes that each hybridize to the two genes involved in the fusion. The test is considered positive if the two separate genes appear close together in signal, but the fusion partner must be known, noting that non-canonical variant fusions may be missed. Amplification can be detected with centromeric enumeration probes, like is done for HER2. By using a probe that targets the centromere of a specific chromosome, you can get information on the chromosomal copy number. Each FISH assay has its own minimum cellularity requirement but, unlike sequencing techniques, the overall percent tumor nuclei in the sample is less of a concern, as long as the tumor nuclei can be easily distinguished from the background normal cellularity. In addition, cytologic slide preparations, such as direct smears, touch imprints, cytospins, and liquid-based preparations allow for examination of the entire nucleus, unlike FFPE sections that transect nuclei and can result in nuclear truncation artifact. As a result, performing FISH on a touch prep slide and preserving the cell block for PDL1 may make a lot of sense in some situations.

Figure 2.

A. Synovial sarcoma, H&E 20X, CT-guided core needle biopsy from left upper lobe of lung. B. Synovial sarcoma; positive SS18 (SYT) paraffin FISH analysis using break apart probes (note numerous isolated red and green signals).

Chromogenic in situ hybridization (CISH)

CISH combines elements of both immunochemistry and FISH. Like FISH, it allows the visualization of specific nucleic acid sequencing in tissue by using a probe; however, like immunochemistry, that probe is visually detected with a chromogen. The EBV encoded RNA (EBER) test is a common CISH used to detect the presence of EBV in tumor cells of patients with nasopharyngeal carcinoma, sinonasal NK cell lymphomas, plasmablastic lymphoma, and some high-grade B-cell lymphomas (Figure 3).^{41, 42} In addition, HR-mRNA ISH is another CISH used most commonly to detect transcriptionally active high-risk HPV in oropharyngeal SqCC tumor cells by hybridizing high-risk specific E6 and E7 mRNA.^{43, 44} CISH assays for HER2 are performed instead of FISH in some labs due to ease of interpretation with a traditional brightfield microscope, precluding a second workflow.⁴⁵

Figure 3.

A. Plasmablastic lymphoma, H&E 50X, CT-guided core needle biopsy of right adnexal mass. B. Plasmablastic lymphoma, Epstein-Barr virus in-situ hybridization stain.

Microsatellite Instability (MSI)

Microsatellite Instability (MSI) testing is a method used to assess the stability of microsatellite repeats in the DNA, which are short (2–10 nucleotides in length), repetitive sequences scattered throughout the genome. MSI testing is often employed to identify tumors with deficiencies in DNA mismatch repair (MMR) mechanisms, which can lead to the accumulation of errors in microsatellite regions that cannot be repaired during DNA synthesis. Two common methodologies for MSI testing are MMR protein immunochemistry and PCR analysis, though NGS assays are also being increasingly utilized.^{18, 46} Four key proteins are typically evaluated to assess the integrity of the MMR complex by immunochemistry. These proteins are MLH1, PMS2, MSH2, and MSH6 (Figure 4). Each protein plays a specific role in the MMR pathway, and their expression levels are examined to identify potential deficiencies in MMR function that may be related to either mutation or promoter methylation. Loss of function in one of these proteins can be identified by a complete lack of nuclear staining by immunochemistry. Based on the findings, additional studies may be warranted to determine if the patient has Lynch syndrome or a sporadic case of MMR deficiency.

Figure 4.

A. Metastatic endometrioid adenocarcinoma of endometrium, H&E 10X, CT-guided core needle biopsy from right lower lobe of lung. B. Metastatic endometrioid adenocarcinoma, MLH1 with intact nuclear expression in tumor cells and background benign cells, 10X. C. Metastatic endometrioid adenocarcinoma, PMS2 with intact nuclear staining in tumor cells and background benign cells, 10X. D. Metastatic endometrioid adenocarcinoma, MSH2 with loss of nuclear staining in tumor cells and positive nuclear staining in background benign cells, 10X. E. Metastatic endometrioid adenocarcinoma, MSH6 with loss of nuclear staining in tumor cells and positive nuclear staining in background benign cells, 10X. Loss of MSH2 and MSH6 in tumor cells corresponds to a deficiency in DNA mismatch repair.

PCR is performed by extracting DNA from the tumor sample and typically adjacent normal tissue. Primers are designed to amplify specific microsatellite loci, and the PCR products are then analyzed for size differences. Microsatellite instability is then identified by the presence of additional or fewer repeats in the tumor compared to the normal sample. A tumor is considered to be MSI-high if there is significant instability in two or more markers, MSI-low if instability in one marker, or microsatellite stable if no instability markers are identified. In summary, MSI testing can be used to help identify individuals at risk for hereditary cancer syndromes and to guide therapeutic strategies, such as selection for immunotherapy, in certain cancer types.

DNA and RNA Sequencing

Although several sequencing methods have been developed over the years, next generation sequencing (NGS) is a versatile, high-throughput, unbiased, parallel technology that can be applied to both DNA and RNA sequencing workflows. As shown in Figure 5, DNA-based NGS assays can detect single nucleotide variants, small insertions and deletions (<~50bp), copy number alterations, and karyotypes, as well as various genomic signatures, such as MSI status, TMB, genomic loss of heterozygosity (LOH), and homologous recombination deficiency (HRD). In contrast, RNA-Seq is the preferred method for detecting gene fusions, largely due to its unique capabilities in directly examining transcripts produced by genes, resulting in increased sensitivity, increased efficiency, and a functionally definitive assessment of gene fusions compared to DNA-based methods.^{47, 48} For example, a recent study including 2,522 NSCLC cases were sequenced with DNA-based methods, and 76.7% were shown to have at least one oncogenic driver alteration.⁴⁷ In a group of 232 cases without driver mutations identified, RNA-Seq was performed and 36 (15.5%) oncogenic fusions were identified that were not detected by DNA-Seq, 33 of which were actionable. One notable strength of RNA-Seq lies in its ability to capture the variability in gene fusion breakpoints at the DNA level. Even when the same fusion arises, different breakpoints within a gene can be identified, providing a more nuanced understanding of genomic alterations. Crucially, RNA-Seq directly identifies fusion transcripts generated by tumor cells only, which improves the limits of detection in cases with low tumor fractions.⁴⁹ Moreover, RNA-Seq’s focus on transcripts helps circumvent the need to interpret DNA variants that do not translate into changes at the RNA level. This emphasis on RNA adds a layer of direct functional relevance to the analysis, ensuring a more comprehensive and accurate assessment of gene fusions and the ability to target them.

Figure 5.

Selected Alterations Detected by DNA and RNA Next Generation Sequencing. A. Single nucleotide polymorphisms, which correspond to variations in a single position of a DNA sequence. These can be silent and result in no change in the amino acid sequence, or can result in a missense mutation (a change in the encoded amino acid) or a nonsense mutation (resulting in a premature stop codon). The pathogenicity depends on additional variables, such as whether or not the change occurs in a noncoding versus coding region, or whether or not the amino acid that is encoded is very biologically different than the wild type amino acid. B. Nucleotide addition, or insertion, is where one or more nucleotides are added to the DNA sequence, thus resulting in a shift in the reading frame or the addition of extraneous codons. C. Internal deletions can also cause a shift in the reading frame, as well as loss of important codons, depending on where in the gene the alteration occurs. D. Gene fusions occur when a segment of DNA moves from one location to another, either within the same chromosome or between chromosome. This change can lead to loss of function of the genes involved or can result in novel fusion products. E. Larger amplifications and deletions can result in partial or complete aneuploidy or polyploidy.

FFPE-based NGS can be performed on small core needle biopsies and cell blocks from FNA and effusion cytology samples that have been fixed in 10% neutral-buffered formalin for between 6 and 72 hours without the use of a strong acid for decalcification, which degrades nucleic acids. Fortunately, biopsies containing bone that have been softened using EDTA chelation agents are typically successful for sequencing.⁵⁰ Specimen requirements are fairly variable depending on the platform and workflow but generally require a tumor volume of 1 mm³ with a tumor nuclei fraction of 20%. When looking at a slide, a 5 mm by 5 mm area equates to 25 mm² of tissue. As long as approximately 10 slides can be cut at 4-micron levels, there should be sufficient tumor volume (1 mm³) to perform most NGS assays. It is important to note that some workflows include tumor microdissection, which can allow for the enrichment of tumor nuclei and drastically improve the nucleic acid extraction process, making previously insufficient samples adequate for sequencing.^{51, 52}

Sequencing can be also be performed on non-FFPE tissue, including from scrapes from FNA smears.⁵¹ The lower limits of detection are dependent on the assay design and platform. For example, the Ion Torrent NGS platform typically requires at least 10 ng of DNA while the Illumina NGS platforms requires a higher input of DNA ranging from 30–270 ng. To put this into context, a single intact diploid cell will yield approximately 6–7 pg of DNA, meaning that a molecular assay that requires 1 ng of DNA would require approximately 1500 intact tumor cells.^53–56

Molecular Diagnostic Testing in Patient Management Decisions

Molecular diagnostic testing on cytology samples plays a crucial role in clinical decision making across various organ systems, including the cervix, thyroid, urine, biliary tract, and pancreatic cyst fluid. Although detailed discussions regarding these site-specific management questions will be addressed in other articles in this series, it is helpful to provide a brief overview here.

HPV Testing in Gynecologic Specimens

The most classic example is the performance of HPV testing, most commonly via PCR, in residual liquid-based Pap preservation media.⁵⁷ The presence of HR-HPV helps determine screening intervals and guide clinical management in cases of indeterminate cytology. In fact, there is a growing push for primary HPV screening, resulting in a reflex cytology evaluation for patients that are shown to be positive for HR-HPV.

Molecular Testing in Thyroid FNA Samples

Next, there are several platforms that assist clinicians in managing patients with indeterminate diagnoses following thyroid FNA. For example, the ThyroSeq^® Genomic Classifier utilizes DNA and RNA NGS to interrogate 112 relevant genes that can parse thyroid lesions into four main classes of molecular alterations.⁵⁸ Although the exact details of the panel are proprietary, the panel can detect mutations in genes such as BRAF, RAS, and TERT, gene fusions such as RET, NTRK, and ALK, copy number alterations like those seen in oncocytic nodules and follicular carcinomas, and gene expression alterations like those seen in medullary thyroid carcinoma or parathyroid nodules. Results are reported as positive or negative with the associated mutations findings, risk of recurrence, and personalized management recommendation. Although the negative predictive value is excellent at >95%, which can help patients avoid an unnecessary surgery, the positive predictive value is only 66%, due to a high rate of false positives from RAS and RAS-like alterations in nodules that end of being benign.^59–63 On the other hand, Veracyte’s Afirma Genomic Sequencing Classifier only utilizes RNA-Seq to analyze expression of 10,196 nuclear and mitochondrial genes, which are used to identify or group nodules into parathyroid tissue, medullary thyroid carcinoma, BRAF V600E mutation, RET::PTC1/PTC3 fusion, follicular cell content index, oncocytic cell index, and oncocytic cell neoplasm index.^{64, 65} Results are reported out as benign (low risk of malignancy, ≤ 4%) or suspicious (high risk of malignancy, ≥ 50%). If the result is suspicious, additional but limited molecular data is reported. In comparison to the ThyroSeq^® panel, Afirma has a similarly high negative predictive value of >95%, but similarly low positive predictive value of approximately 65%.^{64, 66} Additional panels, such as the ThyGeNEXT and ThyraMIRv2, utilizing microRNA expression, exist but with limited clinical applications to date.⁶⁷

Molecular Testing in Urine Samples

Ancillary tests for urinary tract specimens typically aim to achieve one of two objectives. Firstly, ensuring patients under surveillance have no residual or recurrent disease by emphasizing a high negative predictive value (NPV). Secondly, accurately identifying high-grade lesions in situations where biopsy options are limited, requiring a high positive predictive value. Commercial tests often prioritize maximizing NPV, as it impacts a larger patient population.⁶⁸ Ancillary tests for urinary tract specimens come in two types: slide-based and slide-free. Slide-based tests involve a separate slide preparation, potentially impacting lab workflow, while slide-free tests eliminate the need for an additional slide, allowing centralized or on-site testing. Both urinary tract cytology (UTC) and ancillary tests can produce “false false positive” results, where a positive finding occurs without detectable lesions on cystoscopy or biopsy. This scenario can be particularly relevant for certain lesions, like flat CIS, and may prompt more aggressive clinical investigation, potentially enabling earlier detection despite the challenges in assessing new tumors.⁶⁹ Ancillary tests may also uncover dysplastic changes not easily visible with light microscopy. In many circumstances, however, it is difficult to know how to manage a patient with a positive ancillary test and negative cytology/cystoscopy, even knowing with the understanding of the concept of a “false false positive”. Furthermore, these tests are more expensive than urine cytology, labor intensive, and technically difficult. As a result, many of these tests are not widely utilized. See Table 1 for a summary of selected urine ancillary diagnostic tests.

Table 1.

Summary of Selected Ancillary Tests for the Detection of Urothelial Carcinoma in Urine Specimens

Test	Specimen Type	Clinical Utility	Methodology	Sensitivity	Specificity	Additional Comments
UroVysion®	Slide-based	Primary screening and surveillance	Fluorescence in situ hybridization using chromosomal enumeration probes to detect copy number increases in chromosomes 3, 7, and 17 and band assessment of 9p21 to detect homozygous deletion	All Cases −72%70 Non-invasive pTa Tumors −65–73%71, 72 Invasive pT1-T4 Tumors −95–100%71, 72 Upper Tract Tumors −35–87.5%71, 72	All Cases −83%70	-Issue of anticipatory positive: 2/3 of surveillance patients with a positive result and negative cytology and cystoscopy develop recurrence within 29 months73 -Logistical issues with reflex testing, cost, and questionable utility in equivocal cases
uCyt+/ImmunoCyt™	Slide-based	Surveillance	Utilizes three fluorescent antibodies to detect abnormally glycosylated carcinoembryonic antigens	78–90%74, 75	77–87%74, 75	-Protein based test that may result in false positives in setting of urinary tract infections, stones, and etc. -Good sensitivity, especially in detecting low-grade disease
Anti-hTERT	Slide-based	Limited data	Immunochemical stain utilizing an antibody to the catalytic subunit of telomerase	52–60%76	70–91%76	-Not all tumors have alterations in telomerase activity -Non-urothelial cells may stain positive, resulting in challenges with interpretataion
ProEx C™	Slide-based	Limited data	Immunochemical stain utilizing an antibody cocktail to topoisomerase IIα and MMP-2	78%77	96%77	-Likely has low sensitivity with populations with larger numbers of low-grade tumors
NMP22™ BladderChek™	Slide-free	Primary screening and surveillance	ELISA that detects nuclear mitotic apparatus proteins	62–75%75	70–83%75	-Point of care option that results in 30 min -Only requires 4 drops of urine
BTA stat ®, BTA TRAK®	Slide-free	Surveillance	ELISA that detects complement factor H and complement factor H-related protein	54–75%75, 78	64–82%75, 78	-Point of care option that results in 5 minutes -Only requires 5 drops of urine
UroSEEK	Slide-free	Primary screening and surveillance	NGS panel analyzing TERT promoter mutations in addition to a 10-gene panel and aneuploidy analysis	71–95%79	80–93%79	-Theoretical improved performance with multigene and aneuploidy analysis -Decreased sensitivity probably due to increased “false false positives” -Expensive and technically complex

Molecular Testing in Biliary Tract Samples

Biliary tract samples may undergo FISH testing utilizing chromosomal enumeration probes to identify copy number gains in chromosomes 3, 7, and 17, similar to the UroVysion^® test performed in urinary tract cytology. Other labs utilize a FISH panel with probes specific to 1q21, 7q12, 8q24, and 9p21 for polysomy. Depending on the particular FISH assay, sensitivity ranges from 45–55% with a specificity of >95%.⁸⁰ Compared to cytology assessment, which has a sensitivity of around 35% and specificity of 99%, this test offers only a modest improvement in sensitivity.⁸⁰ This test can be performed inhouse or sent to one of several reference laboratories. Samples must be from FNA or biliary brushings submitted in non-gynecologic ThinPrep media or in CytoLyt solution. As an adjunct to conventional cytology assessment, this test may be useful in some settings to increase confidence in a reactive stricture versus a malignant stricture.

Ancillary Testing in FNA Samples of Pancreatic Cyst Fluid

Pancreatic cysts can be challenging to diagnose by FNA.^81–83 The procedure often results in paucicellular cyst fluid contents only without a definitive epithelial component to evaluate. Furthermore, the differential includes a variety of lesions, including reactive pseudocysts, lymphoepithelial cysts, serous cystadenomas, mucinous cystic neoplasms, intraductal papillary mucinous neoplasms, and solid pseudopapillary neoplasm, the last three of which may have dysplasia and may transform to malignancy.⁸⁴ As a result, cell-free ancillary testing can be useful to select patients for surveillance versus surgery. Cyst fluid ancillary testing includes chemistry tests for amylase and CEA analysis, as well as DNA-Seq for KRAS, GNAS, TP53, PIK3CA, PTEN, VHL, and CTNNB1. See the Table 2 below for how this testing can be used to suggest a diagnosis and manage patients.

Table 2.

Summary of Selected Ancillary Tests on Pancreatic Cyst Fluid Specimens

Diagnosis	Amylase	CEA	DNA-Seq	Malignant Potential	Treatment
Pseudocyst	High	Low	None	None	Medical management
Serous cystadenoma	Low	Low	VHL mutation is specific for the diagnosis85	None	Surgery if symptomatic
Mucinous cystic neoplasm	Low	High	KRAS mutation is specific (>90%) but not specific (<50%)86–88 PIK3CA, PTEN, and TP53 may be seen in cases that have transformed to malignancy89–91	Moderate	Surgery
Intraductal papillary mucinous neoplasm	High	High	KRAS and GNAS mutations are specific (>90%) but are not sensitive (<50%)86–88 PIK3CA, PTEN, and TP53 may be seen in cases that have transformed to malignancy89–91	High	Surgery and post-resection surveillance
Solid pseudopapillary mucinous neoplasm	Low	Low	CTNNB1 mutations are specific85	Moderate to high	Resection

Precision Medicine Relies on Predictive Markers

There are over 80 FDA-approved predictive biomarkers with therapy associations. As previously mentioned, these include both tumor specific biomarkers and tumor agnostic biomarkers. The incidence of an actionable alteration is largely dependent on the tumor type. Obtaining the biomarker status of a tumor is recommended for a number of tumor types depending on the clinical scenario, including primary versus metastatic, resectable versus non-resectable, second or third-line therapy assessment, or no other treatment options available. In addition to these scenarios; however, nearly 20% of solid tumors will have one or more tissue agnostic biomarkers, which means preserving tissue for potential future molecular considerations is paramount.^{10, 92, 93} Table 3 includes a selection of tumor specific and tumor agnostic biomarkers, as well as some emerging biomarkers that are likely to be FDA-approved in the near future. It is important to realize that this list is just a portion of a larger group of biomarkers that are only going to increase as our understanding of tumor biology continues and as therapeutic options with proven efficacy continue to be tested.

Table 3.

Selected FDA-Approved and Emerging Tumor Specific and Agnostic Biomarkers

Examples of FDA Approved Tumor Specific Biomarkers	Testing Methodology	Example Tumor Types	Examples of Therapy
ER/PR	IHC	Breast	Tamoxifen, fulvestrant, anastrozole, leuprorelin
HER2	IHC/CISH/FISH	Breast, Gastric/Gastroesophageal Junction	Trastuzumab
PD-L1	IHC	NSCLC, cervical, esophageal, and head and neck SqCC, endocervical adenocarcinoma, urothelial carcinoma, and triple negative breast cancer	Nivolumab, pembrolizumab, atezolizumab, durvalumab
EGFR	DNA-Seq	NSCLC and pancreatic cancer	Erlotinib, osimertinib, gefitinib, afatinib
BRCA	DNA-Seq	Breast, ovarian, fallopian tube, primary peritoneal cancer	Olaparib, rucaparib, niraparib, talazoparib
ALK fusion	FISH/RNA-Seq (IHC screen)	NSCLC	Crizotinib, entrectinib, ceritinib, alectinib, brigatinib, lorlatinib
ROS1 fusion	FISH/RNA-Seq (IHC screen)	NSCLC	Crizotinib, entrectinib, ceritinib, lorlatinib
FOLR1	IHC	Ovarian cancer	mirvetuximab soravtansine
Examples of FDA Approved Tumor Agnostic Biomarkers
MSI-H	PCR/DNA-Seq	Metastatic solid cancers	Pembrolizumab, nivolumab, and nivolumab–ipilimumab combination
dMMR	IHC	Metastatic solid cancers	Pembrolizumab, nivolumab, and nivolumab–ipilimumab combination
TMB	DNA-Seq	Adult and pediatric patients with unresectable or metastatic solid tumors	Pembrolizumab
BRAF V600E	IHC/DNA-Seq	Adult and pediatric unresectable or metastatic solid tumors who have progressed or have no alternative therapies	Dabrafenib with trametinib
RET gene fusion	RNA-Seq	Any solid tumor	Selpercatinib
NTRK fusion	RNA-Seq	Adults and pediatric patients with a solid tumor	Entrectinib, larotrectinib
Examples of Emerging Biomarkers
METex14 skipping	DNA-Seq/RNA-Seq	NSCLC	Capmatinib, tepotinib
FGFR alterations	DNA-Seq/RNA-Seq	Urothelial carcinoma and cholangiocarcinoma	Erdafitinib, pemigatinib
NRG1 fusion	RNA-Seq	NSCLC	Afatinib

Summary:

Over the past decade, the field of cancer diagnostics has seen a remarkable surge in complexity, with a multitude of tumor types and subtypes identified through advanced diagnostic modalities. This evolution has enabled improved risk stratification and treatment options for patients. Notably, the shift towards minimally invasive diagnostic tests, aided by high-resolution imaging and early detection protocols, has allowed for more extensive analysis with smaller tumor samples. Modern cytopathologists, in this intricate landscape, play a pivotal role not just in accurate diagnosis but also as custodians of patient samples, preserving them for relevant biomarker and molecular testing.

Various ancillary testing modalities are utilized for precision medicine in the modern era, including immunochemistry, fluorescence in situ hybridization (FISH), chromogenic in situ hybridization (CISH), microsatellite instability (MSI) testing, and DNA and RNA sequencing. These techniques, each with specific applications and considerations, contribute to the identification of diagnostic, prognostic, and predictive markers critical for clinical decision-making. It is crucial to have a comprehensive understanding of the testing methodology and specimen requirements to navigate the intricacies of sample handling and processing. Already, we have seen the role molecular diagnostic testing impact patient management decisions in areas such as gynecologic, thyroid, urinary tract, biliary tract, and pancreatic cyst fluid samples. Furthermore, small samples from solid tumors from all sites are being increasingly utilized for precision medicine testing. Notably, there are over 80 FDA-approved markers influencing therapy decisions, both tumor-specific and agnostic, marking a crucial paradigm shift in cancer care. Therefore, it is up to the modern cytopathologist to be a great steward in triaging, diagnosing, and handling these specimens.

Key Points:

• Over the past decade, the field of cancer diagnostics has seen a remarkable surge in complexity, with a multitude of tumor types and subtypes identified through advanced diagnostic modalities.

• Minimally invasive diagnostic testing is increasing, meaning modern cytopathologists, beyond their diagnostic role, now safeguard patient samples for vital biomarker and molecular testing, including prognostic and predictive tumor specific and tumor agnostic markers.

• It is crucial for the cytopathologist to have a solid understanding of these ancillary and molecular tests so we can appropriately triage and utilize specimens and adequately and accurately answer clinical care questions.

Clinics Care Points:

• Cytopathologists are tasked with utilizing a number of ancillary testing methods for primary diagnostic, prognostic, and predictive testing, which means specimens must be handled and processed appropriately for a wide range of potential testing.

• Cytopathology samples are commonly used to perform immunochemistry, fluorescence in situ hybridization, chromogenic in situ hybridization, microsatellite instability testing, DNA sequencing, RNA sequencing, and PCR testing, which each have workflow and technical weaknesses and benefits.

• It is important to understand the different molecular diagnostic testing platforms that may be used to inform patient management decisions, including in the gynecologic tract, thyroid, urinary tract, biliary tract, and pancreatic cyst fluid and how to interpret those findings.

• With over 80 FDA-approved predictive biomarkers with therapy, there are increasing indications for ancillary testing beyond making a diagnosis.

• Due to the interdisciplinary complexity in integrating all of this information, some centers are developing molecular tumor boards to improve clinical care.