Is molecular testing of salivary gland FNA specimens ready for prime time?

Abstract

The primary goal of salivary gland fine-needle aspiration (FNA) is facilitating appropriate clinical management, in conjunction with clinical and radiological findings. This includes avoiding surgery mainly for nonneoplastic disease, a subset of benign tumors such as Warthin tumors, and suspected/confirmed lymphomas. On the other hand, the type and grade of salivary gland malignancy is essential to plan the extent of surgery.^{1, 2} In many cases, however, a specific diagnosis is not required. The Milan system for reporting salivary gland cytopathology, which was refined in 2023, has been designated as the current standard reporting scheme in the American Society of Clinical Oncology (ASCO) 2021 guidelines for the management of salivary gland cancer, and risk stratifies salivary gland FNAs into six diagnostic categories with subcategories that are generally sufficient to guide the management including the type and extent of surgery.^{1, 2} Nevertheless, the 2021 ASCO guidelines recommend that pathologists may perform ancillary testing (immunohistochemistry [IHC] or molecular studies) on salivary gland FNAs to support diagnosis and risk of malignancy (ROM).²

Over the last decade, significant advances have been made in the molecular characterization of salivary gland neoplasms.^3-5 Currently, most salivary gland neoplasms including newer entities are characterized by specific genetic alterations, mainly gene fusions that can be used for diagnostic purposes and sometimes therapeutic purposes.^3-5 Furthermore, additional genetic alterations including novel fusions and partners keep being identified, increasing the overall prevalence of genetic alterations across salivary gland neoplasms. Accordingly, there are several ancillary diagnostic tools at our disposal including IHC, fluorescent in situ hybridization (FISH), and more recently next-generation sequencing (NGS) technologies, which can all be successfully applied to FNA specimens, provided there is adequate material to do so.^{1, 6-10} The most common ancillary studies performed on salivary gland FNA include IHC, histochemistry, and FISH, and cell blocks prepared from aspirated salivary gland lesions is the preferred and most convenient material.^6-10 In contrast, NGS is only rarely used by cytopathologists for salivary glands aspirates.^{6, 8, 11-14} Moreover, the choice of the ancillary technique test depends largely on the institution’s budget, available platforms, and expertise of trained staffs. However, NGS is a promising approach to facilitate the diagnosis and classification of salivary gland tumors. Furthermore, the availability and diagnostic yield of NGS techniques have improved significantly over the last decade so that they can be used to confirm the diagnosis of specific entities.^15-18 Briefly, the basic principle of NGS is massive parallel sequencing of specific DNA regions of interest, either through whole exome, transcriptome, or targeted panel sequencing.

New types of RNA-based NGS, such as RNA-Seq have become increasingly used in the detection of fusion transcripts, a common molecular alteration seen not only in lymphomas and sarcomas but also salivary gland tumors.^{5, 13-15, 18} The advantages of NGS and specifically RNA-Seq, compared to other methodologies such as FISH, include broader mutation coverage, the interrogation of 100 to over 1000 potential fusions simultaneously, covering most of the common molecular alterations of salivary gland tumors in a single test.^{5, 15} Moreover, whole transcriptome RNA-Seq or targeted RNA sequencing using hybridization-based or anchored multiplex polymerase chain reaction approaches can lead to the possible identification of novel fusion partners.^{5, 13, 14} Currently, there are multiple commercially available fusion assays on the market including TruSight RNA Fusion Panel and TruSight RNA Pan-Cancer Panel targeting 507 and 1385 known fusion-related genes, respectively, using the Illumina MiSeq system. From a practical angle, targeted fusion panels allow for the routine use of formalin-fixed paraffin-embedded (FFPE) tissue, allowing for easier integration of this method into the workflow of a surgical pathology laboratory.⁵ However, despite these benefits, the disadvantages are that RNA-Seq remains an expensive and time-consuming tool. Therefore, NGS does not yet have a routine diagnostic role in salivary gland cytopathology as it is not widely available in many centers/laboratories.⁶ Furthermore, there are now several commercially available IHC markers, including PLAG-1, MYB, pan-TRK, and NR4A3, that can serve as surrogates for underlying molecular alterations, contributing to the accurate diagnosis and representing a convenient and inexpensive alternative to NGS.^{1, 6-10}

Over the last few years, some institutions have developed their own comprehensive customizable NGS with broad panels specifically designed to cover most gene alterations, including mutations, fusions ,and RNA gene expression alterations in salivary gland tumors.^{3, 12, 14} Among them, the SalvGlandDx panel is an all-in-one RNA-based NGS panel for the detection of mutations, fusions and gene expression levels of 27 genes involved in salivary gland tumors.^{3, 14} The SalvGlandDx panel covers most of the common molecular alterations of salivary gland tumors in a single test and can be reliably performed on FFPE cell block specimens.^{3, 14}

In their exploratory multicenter study, Freiberger et al.¹⁹ show that comprehensive RNA-based molecular sequencing (including gene fusions, gene mutations and altered gene expression) of both cytological smears and cell block samples of salivary gland FNA specimens using the SalvGlandDx panel is technically feasible, with the smear preparations showing a higher success rate (62.5% vs. 41.1%) in their feasibility cohort. Sequencing was successfully performed at three different centers using either Illumina or ThermoFisher based platforms, revealing frequent molecular alterations in Milan IVB and Milan V cases. Furthermore, their results were promising for improving the classification of indeterminate SUMP (Milan IVB) cases into clearly benign (Milan IVA) or malignant (Milan VI) categories.

A total of 106 cytological cases were evaluated in this study. Of 63 technically valid cases, 76% revealed a molecular alteration and 94% of these molecularly altered cases could be assigned to a different Milan category when additionally taking molecular results into account. In only 2% of the SUMP cases in which a molecular alteration was detected, the classification remained unchanged. The authors evaluated the ROM of the cytological specimens considering the molecular findings. For PLAG1 and HMGA2 gene fusions, the ROM was 3.4%, which is in keeping with the expected ROM of Milan IVA (Neoplasm: Benign) category.¹ The only malignant case in their series was a salivary duct carcinoma ex pleomorphic adenoma (PA) with HMGA2 rearrangement. However, the case demonstrated cytologic atypia, and was therefore morphologically classified as Milan V. Although nonspecific for PA, the PLAG1-positive cases were most frequently found in PA. These data further corroborate the findings from Sanchez-Avila et al.⁹ that, in the absence of atypical features, PLAG1 rearrangements are most frequently found in the spectrum of cellular PA. This suggests that finding PLAG1 in a basaloid SUMP (by IHC and/or NGS) may allow for a reclassification into Milan IVA category.^{9, 19} PLAG1 fusions have been also described in approximately 40% of myoepitheliomas and myoepithelial carcinomas, showing the molecular similarities of these tumors and emphasizing the importance of molecular-morphological correlation.^{3, 4, 19}

Although only a few cases of MAML2, ETV6, and MYB rearranged neoplasms were included in their study, the ROM associated with these alterations was 100%.¹⁹ Regarding gene mutations, the most common one in their series was HRAS, and all cases were subsequently diagnosed as malignant in the resection specimens, including epithelial-myoepithelial carcinoma, salivary duct carcinoma, or intraductal carcinoma.¹⁸ Thus, the ROM for HRAS mutations in their cohort was 100%. In contrast, CTNNB1 mutations were found in basal cell adenoma in their cohort, corresponding to a ROM of 0%.

Interestingly, only five cases of Milan category III (atypia of undetermined significance) were included in the study, and the single analyzable case by NGS, corresponding to nodular oncocytic hyperplasia as final diagnosis, revealed no molecular alteration.¹⁹ This highlights the difficulty of performing successful ancillary studies in this indeterminate diagnostic category, in which samples are often compromised by the scant cellularity, and supporting the role of a repeat FNA for these challenging cases.¹

In summary, the study by Freiberger et al.¹⁹ contributes to characterizing the molecular spectrum of indeterminate lesions in salivary gland FNA. However, it represents only a limited selection of challenging/indeterminate cases and additional studies are necessary to validate these findings and especially when determining the ROM for certain genetic alterations (e.g., PLAG1). Except for PLAG1 and HMGA2 alterations that are frequently retrieved in PAs, gene fusions in salivary gland neoplasms are associated with malignant tumors and are often specific for a given pathological entity.^{3, 4} On the other hand, point mutations such as HRAS are less common and less specific in general but can also be very helpful to further classify challenging FNAs when integrated with cytomorphology and clinico-radiological features. Interestingly, there are some similitudes with molecular testing of indeterminate thyroid FNAs where each molecular alteration may be associated with a well-defined ROM allowing refining the diagnostic category and/or guiding management.²⁰ For example, BRAF V600E mutation is associated with nearly 100% ROM, whereas RAS mutation is nonspecific and may be associated with the whole spectrum of follicular neoplasms.²⁰

With the improved molecular characterization of salivary gland neoplasms, FNA specimens including smears and cell blocks are amenable to molecular testing for defining alterations, providing a specific diagnosis in many cases and helping to guide the management in preoperative setting. Molecular testing can be particularly beneficial and complimentary when used within the framework of the Milan system as the results can further refine the ROM, especially in indeterminate categories IVB and V, often allowing the definitive histotyping of salivary gland neoplasms, as recommended by the recent ASCO guidelines.² For practical purposes, NGS is most useful for confirming a specific diagnosis that is strongly favored by the cytomorphological, immunohistochemical, and/or clinico-radiological features. However, although this is conceptually very attractive, data establishing performance characteristics for testing on FNA are still limited to small series and case reports. Further studies with larger cohorts are required to evaluate the comprehensive role of NGS on salivary gland FNA specimens. It appears that molecular testing of salivary gland FNA is at least 1 decade behind thyroid FNA. Given the overall success of molecular testing for indeterminate thyroid FNAs, which is now well endorsed in the guidelines, however, there is no reason to believe that a similar fate would not be achievable for salivary gland FNA in the near future. Molecular testing of salivary gland FNA using NGS will likely play a larger role for diagnostic purposes in the future, as the technology keeps improving while becoming cheaper and more widely available. However, it will likely take much longer for thyroid to fully endorse molecular testing of salivary gland FNA using NGS given the rarity of salivary gland tumors in comparison with thyroid. Thus, although molecular testing of salivary gland FNA specimens may not be ready for prime time just yet, we have reached a significant milestone. Akin to thyroid, it is also likely that the future of salivary gland molecular cytopathology will involve not only diagnostics but will also be critical for informing prognosis and therapeutics, especially in patients with advanced nonresecable and/or metastatic salivary gland tumors.

The author declares no conflicts of interest.