CDKN2A/p16 evaluation in cytology specimens.

Cyclin-dependent kinase inhibitor 2A (CDKN2A) is an important tumor suppressor gene located at 9p21.3 that expresses p16 and p19 through alternate reading frames. Mutations or losses of CDKN2A are central to tumorigenesis in a wide variety of tumors because p16 plays an important role in cell cycle regulation through inhibition of CDK4 and CDK6. p16 activity on CDK4/6 reduces phosphorylation of pRB and thus reduces transcriptional activity of E2F and arrests cell cycle progression from the G1 phase to the S phase (Figure 1A). p19 interacts with MDM2, and this results in p53 stabilization and promotes cell cycle arrest and apoptosis. An evaluation of CDKN2A/p16 status and/or expression can be useful in cytology specimens in a few specific circumstances, as detailed in this brief article. Diagrammatic representation of p16 function (A) in normal cell cycle regulation, and (B) in cancers with a loss of p16 function and (C) upregulation of p16 in HPV-associated cancers. HPV indicates human papillomavirus. Approximately 50% of all human cancers show some form of p16 inactivation. Reduced p16 function is tumorigenic because of increased cell cycling from reduced negative regulation on the G1–S checkpoint (Figure 1B). A reduction or loss of p16 function occurs through a variety of mechanisms, including point mutations, loss of heterozygosity, homozygous deletion (HD), and promoter methylation. The precise mechanisms and their proportions are dependent on the tumor type.1-5 Interestingly, a loss of p16 function correlates with more aggressive disease in some tumor types.6-10 Despite the frequency of altered p16 function in cancer, molecular evaluation of p16 alterations is not frequently used clinically in cytology samples. One exception to this is the increasing use of p16 evaluation for effusion cytology specimens for mesothelioma. In addition, immunohistochemical detection of p16 overexpression is an important diagnostic tool in the diagnosis of high-risk human papillomavirus (HR-HPV)–associated oropharyngeal squamous cell carcinomas (OPSqCCs) from cytology specimens. Malignant pleural mesothelioma (MPM) is an aggressive malignancy, and patients often present with pleural effusions. Although pleural biopsy is the gold standard for diagnosis, this procedure is invasive and may not be tolerated. Additionally, effusion cytology specimens are usually received in pathology laboratories earlier in a patient’s course, and this potentially allows an earlier diagnosis. Making an accurate and prompt diagnosis of MPM with cytology specimens is paramount. An increasing number of biomarkers are now available to distinguish reactive mesothelial proliferations from mesothelioma. Previously, the International Mesothelioma Interest Group stated that cytology was insufficient for a diagnosis of MPM; however, more recent guidelines include cytology as an acceptable method but require the incorporation of ancillary studies for confirmation.11 In the following discussion, it is also important to note that the high specificity of these studies is dependent on confirmation of the mesothelial nature of tumor cells (i.e., not adenocarcinoma). CDKN2A/p16 HD is the most common genetic alteration in MPM and is a useful diagnostic marker for distinguishing reactive mesothelial cells from MPM in cytology specimens.12 A meta-analysis of published data has shown that CDKN2A/p16 HD, assessed by fluorescence in situ hybridization (FISH), has 100% specificity and a moderate sensitivity of 62% for the diagnosis of MPM.13CDKN2A/p16 HD is seen at similar rates regardless of the histologic subtype of MPM (approximately 70%), but it appears to be less common in peritoneal mesothelioma (approximately 25%).14CDKN2A/p16 HD also has prognostic value, and CDKN2A/p16 HD portends worse overall survival for patients.15CDKN2A/p16 FISH can be performed on cell blocks or smears.16 p16 immunohistochemistry has no role in the evaluation for MPM, as protein expression does not correlate well with gene deletion. Methylthioadenosine phosphorylase (MTAP) is frequently codeleted with CDKN2A, as these genes are in close proximity on chromosome 9. MTAP immunohistochemistry demonstrates cytoplasmic loss of expression in many cases with CDKN2A HD and has been proposed as a surrogate for CDKN2A FISH.17 MTAP immunohistochemistry may offer a cheap and fast alternative to CDKN2A FISH but is currently less well studied in effusion specimens, and CDKN2A FISH appears to be slightly more sensitive.18, 19 Loss of nuclear BRCA1-associated protein 1 (BAP1) expression by immunohistochemistry is also useful in effusion samples with atypical mesothelial cells.20 BAP1 loss has essentially 100% specificity but only moderate sensitivity for epithelioid MPM.21 Although not frequently encountered in effusion specimens, sarcomatoid MPM has lower rates of BAP1 loss; this results in potentially poor sensitivity in such circumstances. BAP1 may also be less reliable in alcohol-fixed specimens because of antigen leaching.18 Although each of these ancillary tests can help to confirm a diagnosis of MPM, BAP1 immunohistochemistry followed by an examination for a 9p21 deletion with CDKN2A FISH or MTAP immunohistochemistry if BAP1 is retained is recommended by many authors.13, 18 Notably, the loss of these markers is also seen with mesothelioma in situ. Although this offers the ability to diagnose early even mesothelioma in situ, the inability to identify invasive disease in most cytologic specimens means that a distinction between invasive MPM and mesothelioma in situ is not always possible. Additionally, the evaluation of these assays (particularly immunostains) can be challenging on cytologic samples, with issues including inadequate staining of positive controls, a lack of an internal positive control for cytology specimens, heterogeneous staining of tumor cells, and inadequate tumor cellularity for evaluation.22 Transcriptionally active HR-HPV infection leads to the expression of viral oncogene E6 and E7 proteins, and this subsequently results in inhibition of p53 and pRB (Figure 1C). The inhibitory effect of E7 on pRB results in compensatory cellular upregulation of p16. As a result, p16 detection by immunohistochemistry is an excellent surrogate marker for HR-HPV but only in particular instances. Within the head and neck, the majority of oropharyngeal carcinomas are associated with HR-HPV defining the diagnosis; therefore, confirmation of HPV status is required for squamous cell carcinoma (SqCC) of the oropharynx. In the uterine cervix, the vast majority of carcinomas are related to HR-HPV, and its detection is helpful in screening. HR-HPV–associated OPSqCC has become increasingly common.23 Human papillomavirus (HPV)–associated OPSqCC tends to affect younger White men and is more closely associated with sexual risk factors than tobacco use. HPV-associated OPSqCC has a better prognosis and response to therapy than conventional SqCC of the head and neck, and determination of the HR-HPV status is therefore critical for its diagnosis and management.24 Furthermore, diagnostic material is often first obtained by fine-needle aspiration of a metastatic neck lymph node, and these cytology samples may be the only specimens available for HPV testing. College of American Pathologists (CAP)25 and American Society of Clinical Oncology (ASCO)26 guidelines indicate when HPV testing is appropriate for head and neck SqCC. These guidelines agree that all oropharyngeal primaries and metastatic SqCCs of an unknown primary must be tested. Such testing is also required for proper staging according to the eighth edition of the American Joint Commission on Cancer staging manual.27 CAP and ASCO allow for p16 immunohistochemistry alone to be used as a surrogate marker when a patient has a known oropharyngeal primary. However, ASCO indicates the need for direct HPV testing for metastases of an unknown primary, whereas CAP allows for p16 alone for metastases to cervical levels 2 and 3 with nonkeratinizing morphology.28 There are several important caveats in the consideration of p16 for determining HPV status. p16 immunohistochemistry serves as a surrogate marker for HR-HPV in the setting of OPSqCC only because of the high prevalence of transcriptionally active HR-HPV in these tumors. However, p16 can be positive in any tumor regardless of the HPV status; thus, any setting in which HPV is less common will result in a lower positive predictive value.29 For example, HR-HPV also causes SqCC in various other subsites in the head and neck but at a much lower rate in comparison with the oropharynx. Specifically, a significant subset of sinonasal SqCCs (the second highest subsite) are associated with HR-HPV, but even there, p16 can serve only to screen for transcriptionally active HR-HPV; a direct testing method is needed.30 Similarly, although SqCC metastases of an unknown primary to cervical levels 2 and 3 are likely to be from an oropharyngeal primary, the uncertainty of this lowers the accuracy of p16 in predicting the HPV status. Although CAP guidelines allow p16 testing in this setting, ASCO recommends direct testing of metastases of an unknown primary. Even in known oropharyngeal tumors, there is a known rate of p16/HPV discrepancy (approximately 10% of tumors), and this has been shown to have prognostic significance.31 More direct testing methods may be recommended in future guidelines.28 Although p16/HPV testing of aspirate samples is quite accurate, several challenges, including variable cellularity, fixation, and preparation, exist. Multiple studies suggest that the threshold for p16 on cell blocks should be lower than the 70% cutoff used for histology. Additionally, alcohol fixation can reduce antigenicity in comparison with formalin-fixed cell blocks, and our experience shows this to be the case with p16.32-34 Regardless of the cell block process used in one’s laboratory, a lower p16 threshold for cytology specimens is likely needed, and each laboratory should perform an independent validation for the appropriate cutoff used.35 Direct HR-HPV testing methods, including in situ hybridization and polymerase chain reaction, can also be performed on cytology specimens. Additionally, aspirate samples collected in liquid-based cytology media can be validated with the laboratory’s method for testing cervicovaginal (Papanicolaou) tests.36-39 Such methods have been shown to be highly accurate and can detect HPV even in cystic, hypocellular samples. Such testing should be internally validated if it is to be used clinically. In summary, despite the common molecular alterations seen in CDKN2A/p16 in many cancer types, there are only particular settings in which routine evaluation is clinically useful and relevant. Identification of CDKN2A/p16 HD confirms the diagnosis of MPM in effusion specimens containing atypical mesothelial cells. p16 upregulation is also an important finding for determining the HR-HPV status in OPSqCC, but this can be further confirmed with direct HPV testing. Despite the importance and usefulness of testing in these settings, there are important caveats that must be considered. Christopher C. Griffith reports being an independent contractor for the United States and Canadian Academy of Pathology and Scientific Symposiums International. The other author declares no conflicts of interest. Juan Xing is a staff pathologist in the Department of Pathology at the Cleveland Clinic’s Pathology & Laboratory Medicine Institute. She practices cytopathology. Her research interests are primarily in the areas of nongynecological cytopathology, telecytology, and the evaluation of a variety of ancillary tests in cytological specimens. Christopher C. Griffith is a staff pathologist at the Cleveland Clinic in Cleveland, Ohio. He is the director of the Head and Neck Pathology Service and the director of the Head and Neck Surgical Pathology Fellowship. His clinical and research interests include work to improve the accuracy of diagnosis in aspirates from adults with neck cysts. This work has primarily focused on the importance of determining the human papillomavirus (HPV) status and methods for HPV testing in aspirate material.