The Diagnostic Utility of RAS Q61R Mutation-specific Immunohistochemistry in Epithelial-Myoepithelial Carcinoma

Abstract

Epithelial-myoepithelial carcinoma (EMC) is a rare salivary gland cancer characterized by biphasic tubular structures composed of inner ductal and outer clear myoepithelial cells. Because of its histological variety and overlap of histological features with other salivary gland tumors, there are broad differential diagnoses. The HRAS Q61R mutation has been reported to be frequent in and specific to EMC. We evaluated the usefulness of RAS Q61R mutant-specific immunohistochemical staining for detecting this genetic alteration in EMC. We investigated 83 EMC cases and 66 cases of salivary gland tumors with an EMC-like component, including pleomorphic adenoma, adenoid cystic carcinoma, basal cell adenoma/adenocarcinoma, and myoepithelial carcinoma. Sanger sequencing was performed for HRAS, KRAS, and NRAS. The diffuse and membranous/cytoplasmic RAS Q61R immunohistochemical expression was observed in 65% of EMC cases, in which all cases harbored the HRAS Q61R mutation. Immunohistochemistry-positive cases were present only in de novo EMCs (54/76 cases, 71%) but not in EMCs ex pleomorphic adenoma. The immunoreactivity was almost always restricted to the myoepithelial cells. Conversely, all EMC cases lacking the HRAS Q61R mutation were negative on immunohistochemistry. In addition, only 3% of EMC-like tumors showed the abovementioned immunopositivity. None of the cases examined carried KRAS or NRAS mutations. Immunohistochemistry for RAS Q61R is highly sensitive and specific for detecting the HRAS Q61R mutation in EMC. Since significant immunopositivity was almost exclusively identified in nearly two-thirds of EMCs but seldom in the histologic mimics, the immunohistochemistry of RAS Q61R is a useful tool for diagnosing EMC in general pathology laboratories.

INTRODUCTION

Epithelial-myoepithelial carcinoma (EMC) is a rare low-grade salivary gland carcinoma that is histologically characterized by biphasic tubular structures composed of inner ductal and outer clear myoepithelial cells.¹ However, in addition to this classical features, EMC exhibits several histologic variations resembling other tumor entities.^2–6 EMC usually arises de novo but uncommonly occurs via evolution from a preexisting pleomorphic adenoma (PA) (i.e. carcinoma ex PA).^{2, 3} Furthermore, similar to EMC, many other benign and malignant salivary gland tumor types, such as PA, adenoid cystic carcinoma (AdCC), and basal cell adenoma/adenocarcinoma, also show biphasic differentiation, and they sometimes have focal clear myoepithelial components. Myoepithelial salivary gland tumors can also display clear cell morphologies and share histological findings of myoepithelial-predominant EMC. Therefore, making an accurate diagnosis of EMC may be quite challenging for general surgical pathologists, especially when only small tissue fragments are available for a histologic examination.

Recently, frequent and specific HRAS mutations have been reported in EMC, particularly in de novo EMC.^{2, 7, 8} HRAS is a member of the RAS family of proteins (HRAS, NRAS, and KRAS) that reside at the cytosolic side of the plasma membrane and transmit proliferation signals through the intracellular RAS-RAF-MEK-MAPK pathway.^{9, 10} The occurrence rate of HRAS mutations in EMC is said to range from 26.7% to 81.7%, and the most common mutation locus is at codon 61, causing a glutamine to be replaced with an arginine (Q61R), followed by other mutation patterns, including G13R and Q61K.^{2, 7, 8} In contrast, no HRAS mutations have been detected in other benign or malignant salivary gland tumors that have EMC-like clear myoepithelial cells.² The identification of an HRAS mutation is thus a powerful diagnostic tool for EMC.²

In the era of next-generation sequencing, a mutation analysis of a specific gene is a simple and low-cost test; however, molecular techniques are still not widely available outside of academic medical centers, and they can be time-consuming compared to immunohistochemistry (IHC). Recently, IHC for several mutant-specific proteins, including BRAF V600E and IDH1 R132H, has been widely used in clinical practice as a surrogate technique for a mutation analysis.^{11, 12} The RAS Q61R mutation-specific antibody was originally intended to detect NRAS Q61R mutant protein,^{13, 14} but recent studies have shown that it is also reactive for HRAS Q61R and KRAS Q61R mutant protein with high sensitivity and specificity, as these RAS family proteins share almost 100% sequence homology of the amino-terminal catalytic domain (amino acids 1–165).^15–18 To our knowledge, the diagnostic significance of IHC for RAS Q61R in EMCs has not yet been described.

In this study, we performed RAS Q61R IHC and compared the results with HRAS mutation status in a large series of EMCs and salivary gland tumor mimics to determine the usefulness of RAS Q61R IHC as an alternative marker of the HRAS Q61R mutation in the differential diagnosis of EMC.

MATERIAL AND METHODS

Patients and Histological Review

The present study was approved by the Institutional Ethics Review Board of the ethics committee of each collaborating institution, and the need to obtain informed consent was waived due to the retrospective nature of the analysis. Three board-certified pathologists (M.N., M.U., and N.T.) carefully reviewed all cases according to the criteria of the 2017 WHO Classification of Head and Neck Tumours.¹ Histologic variations of EMC were recorded regardless of the proportion of the tumor. Classic EMC was defined as a de novo EMC lacking any features of specific morphologic variations.²

In total, 105 EMC cases in the institutional and consultation files were retrieved, among which 83 cases with available unstained slides were included in the present study cohort and subjected to a mutation analysis and RAS Q61R IHC. Among them, 62 cases had been included in the previous mutation analysis study cohort.² For the comparison, we also strictly selected other salivary gland tumors partially but closely resembling EMC with clear myoepithelial cells; these tumor types included AdCC (n=21), PA (n=15), basal cell adenoma/adenocarcinoma (n=18), and myoepithelial carcinoma (n=12).

Mutation Analyses

Polymerase chain reaction and subsequent Sanger sequencing were performed for DNA extracted from unstained formalin-fixed paraffin-embedded slides. The tumor component of the slides was microdissected to increase the tumor cell ratio. The protocol was the same as that described in the previous study.² A hotspot mutation analysis of HRAS (exons 2 and 3) and KRAS (exons 2 and 3) was performed for all analyzable cases. An NRAS (exons 2 and 3) mutation analysis was only performed for the few cases with discordant findings between the HRAS mutation analysis and RAS Q61R IHC. The primer sequences are listed in Supplementary Table S1 (Supplemental Digital Content 1). Mutation analyses were performed after a histological diagnosis was established.

Immunohistochemistry

Immunohistochemical studies were performed on 4-μm-thick formalin-fixed paraffin-embedded sections. The antibodies used to aid in the histological diagnosis of EMC are listed in supplementary Table S2 (Supplemental Digital Content 2). RAS Q61R mutation-specific IHC was conducted after the diagnosis of each tumor with the monoclonal antibody specific to the NRAS Q61R mutant (clone SP174; Abcam, Cambridge, MA) at a dilution of 1:100. This recombinant antibody also recognizes Q61R-mutated HRAS and KRAS as well as NRAS.^13–17 Heat-induced antigen retrieval was performed with an autoclave (105 °C, 20 minutes) in antigen retrieval buffer (Nichirei Bioscience, Tokyo, Japan). N-Histofine Simple Stain Max PO (Nichirei Bioscience) was used as a secondary reagent. The results of RAS Q61R IHC were evaluated by three board-certified pathologists (M.N., M.T. and T.N.) independently from each other and while blinded to the results of the mutation analysis.

Staining of RAS Q61R IHC was scored of 0 to 3+: 0, no membranous or cytoplasmic reactivity; 1+, diffuse faint to barely perceptible reactivity or irregular focal reactivity; 2+, weak to moderate diffuse membranous or cytoplasmic staining in the tumor; and 3+, moderate to intense diffuse membranous or cytoplasmic reactivity. Nuclear staining was regarded as a non-specific signal and ignored, regardless of its positivity. Cases with scores of 2+ and 3+ were considered positive, and those with scores of 0 and 1+ were considered negative.

RESULTS

Clinical and Histological Characteristics

The clinical characteristics of the current cohort are summarized in Table 1. The patients’ ages ranged from 35 to 93 (mean, 69.7) years old, with a female:male ratio of 43:38. Tumors ranged in size from 0.8 to 7.5 cm (mean, 3.5 cm). The most common anatomical site was the parotid gland, comprising 80% of the cases (66/83 cases), followed by the submandibular gland, bronchus, nasal cavity, oral cavity, and other minor salivary glands. A total of 92% (76/83 cases) were de novo EMC, while 8.4% (7/83 cases) arose from a preexisting PA. The clinical features of other tumor types are shown in Supplementary Table S3 (Supplemental Digital Content 3).

TABLE 1.

Clinicopathological Characteristics of Epithelial-Myoepithelial Carcinoma Patients (n=83)

Age (mean [range]) (years)	69.7 (35–93)
Sex (n [%])
Female	43 (52)
Male	38 (46)
Unknown	2 (2.4)
Size (mean [range]) (cm)	3.5 (0.8–7.5)
Anatomic site (n [%])
Parotid gland	66 (80)
Submandibular gland	5 (6.0)
Bronchus	4 (4.8)
Nasal cavity	3 (3.6)
Oral cavity	2 (2.4)
Parapharyngeal space	1 (1.2)
Ethmoid sinus	1 (1.2)
Unknown	1 (1.2)
Histologic origin
De novo	76 (92)
Ex pleomorphic adenoma	7 (8.4)
Histologic features
Classic	28 (34)
Non-classic	55 (66)
Cribriform	16 (19)
Basaloid	15 (18)
Sebaceous	13 (16)
Papillary-cystic	13 (16)
Oncocytic/Apocrine	9 (11)
Psammomatous	3 (3.6)
Double-clear	2 (2.4)
Squamous	1 (1.2)
Verocay-like	1 (1.2)
High-grade transformation	3 (3.6)

Histologically, 66% of the cohort (55/83 cases) exhibited histological variations, some of which overlapped, and the remainder of cases were classified as classic EMC (Table 1). The histomorphologic details of the variations were described in the previous study.²

Results of Mutation Analyses

The results are summarized in Table 2. The majority of EMC cases harbored HRAS mutations (66/83 cases [80%]), most of which were Q61R (55/66 cases [83%]), with the remainder including G13R and Q61K. Concerning the histological origin, HRAS mutations were identified only in de novo EMCs (66/76 cases [87%]) and not in EMCs ex PA (0/7 cases [0%]). No KRAS or NRAS mutations were identified in the EMC cases examined. All EMC-like salivary gland tumors lacked hotspot mutations in HRAS, KRAS, and NRAS.

TABLE 2.

The HRAS Mutation Status in Epithelial-Myoepithelial Carcinoma and Its Mimics (n/N [%])

	HRAS mutation status	Total	De novo	Ex pleomorphic adenoma
EMC	Wild type	17/83 (20)	10/76 (13)	7/7 (100)
	Mutated	66/83 (80)	66/76 (87)	0/7 (0)
	G13R	4	4	0
	Q61R	55	55	0
	Q61K	7	7	0
EMC mimics	Wild type	66/66 (100)	66/66 (100)	0 (0)

EMC indicates epithelial-myoepithelial carcinoma.

RAS Q61R Immunohistochemistry

The results of RAS Q61R IHC are summarized in Table 3. Representative examples of RAS Q61R IHC are shown in Figs. 1–4.

TABLE 3.

Results of RAS Q61R Immunohistochemistry in Epithelial-Myoepithelial Carcinoma and Its Mimics

	Positive			Negative		Total	Positive ratio (%)
	3+	2+		1+	0
EMC	49	5		2	27	83	65
De novo	49	5		1	21	76	71
Ex pleomorphic adenoma	0	0		1	6	7	0
EMC mimics	2	0		2	62	66	3.0
Adenoid cystic carcinoma	0	0		1	20	21	0
Pleomorphic adenoma	1	0		1	13	15	6.7
Basal cell adenoma/adenocarcinoma	1	0		0	17	18	5.6
Myoepithelial carcinoma	0	0		0	12	12	0

EMC indicates epithelial-myoepithelial carcinoma.

FIGURE 1.

The histological, immunohistochemical, and genetic features of epithelial-myoepithelial carcinoma (EMC). (A-D) Classical EMC with an HRAS Q61R mutation. A, The tumor exhibits bi-layered tubular structures consisting of inner eosinophilic ductal cells and outer polygonal clear myoepithelial cells. B, The immunoreactivity for RAS Q61R is diffusely observed in the tumor. C, High-power view reveals that the cytoplasmic and membranous immunostaining is restricted to the outer myoepithelial cells of EMC. Note that the inner luminal ductal cells are devoid of the expression. D, Sanger sequencing showing an HRAS Q61R mutation. (E, F) Myoepithelial-predominant EMC. E, The tumor is mostly composed of clear myoepithelial cells. F, The tumor cells are diffusely positive for RAS Q61R immunohistochemistry, whereas ductal cells are indistinguishable in this area. (G, H) EMC with an HRAS Q61K mutation. G, The tumor cells are completely negative for RAS Q61R immunohistochemistry. H, Sanger sequencing of the same case as (G).

In this study, given the protein localization of RAS family protein, since only the membranous and cytoplasmic expression is meaningful and suitable for the evaluation, nuclear staining was regarded as non-specific immunoreactivity. Generally, the positive expression of RAS Q61R was diffuse with finely granular cytoplasmic and linear membranous patterns, and it was easily interpreted. Background normal salivary tissue showed no signal or a very faint signal. Differences in positivity between tumor cells and the surrounding normal salivary tissue were evident in cases showing a score of 2+ or 3+, so these groups were considered positive. A score of 1+, which was considered negative, was assigned when the expression was so weak that it was almost indiscernible.

A total of 65% of EMCs (54/83 cases) were positive for RAS Q61R IHC, most of which were scored as 3+ (Table 3, Figs. 1A, B). IHC-positive cases were present only in de novo EMCs (54/76 cases, 71%), and all EMC ex PA cases were negative. The immunoreactivity was restricted to the myoepithelial cells, and the luminal cells were usually negative. In the cases with apparent biphasic differentiation, the two cell populations were easily distinguished by the difference in the RAS Q61R expression pattern (Fig. 1C). However, in the myoepithelial-predominant area or tumors, the luminal cells were limited and difficult to recognize (Figs. 1E, F).

The correlation between the RAS Q61R IHC and HRAS mutation status finding is shown in Table 4. Almost all cases of EMC harboring an HRAS Q61R mutation were RAS Q61R IHC-positive (54/55 cases [98%]) (Figs. 1A–D). The remaining case showed very weak expression and was scored as 1+. Conversely, EMC cases that were HRAS wild-type (17 cases) and those with HRAS mutations other than Q61R, including Q61K (7 cases) and G13R (4 cases), were all negative for RAS Q61R IHC (Figs. 1G, H).

TABLE 4.

Correlation between the HRAS Mutation Status and RAS Q61R Immunoreactivity in Epithelial-Myoepithelial Carcinoma

HRAS mutation status	RAS Q61R immunohistochemistry
	Positive			Negative		Total	Positive ratio (%)
	3+	2+		1+	0
Q61R	49	5		1	0	55	98.2
Q61K	0	0		0	7	7	0
G13R	0	0		0	4	4	0
Wild-type	0	0		1	16	17	0

EMCs exhibiting histological variations, such as cribriform structure (Figs. 2A, B), a basaloid appearance, apocrine differentiation (Figs. 2C, D), papillary-cystic architecture, and squamous differentiation (Figs. 2E, F), demonstrated diffuse immunoreactivity with a myoepithelial cell-specific staining pattern, which was identical to the pattern seen in the classic form of EMC (Supplementary Table S4, Supplemental Digital Content 4). In addition, one of three EMC cases with high-grade transformation was also positive for RAS Q61R IHC. In this case, both the classic EMC and high-grade carcinoma components diffusely expressed RAS Q61R. Of note, the luminal cells with sebaceous differentiation showed strong signals regardless of the HRAS mutation status (Fig. 3). Consequently, in the sebaceous EMC, we carefully judged a case to be negative when the expression was restricted to the sebaceous luminal cells and was absent in the myoepithelial cells (Figs. 3C, D).

FIGURE 2.

Combination of the histological and RAS Q61R immunohistochemical features of epithelial-myoepithelial carcinoma exhibiting histological variations. (A, B) Cribriform structure. (C, D) Apocrine differentiation with papillary growth pattern. (E, F) Squamous differentiation.

FIGURE 3.

The histological and immunohistochemical features of epithelial-myoepithelial carcinoma (EMC) showing sebaceous differentiation with or without an HRAS Q61R mutation. (A, B) HRAS Q61R-mutated sebaceous EMC. B, Strong immunoreactivity is observed in both the inner sebaceous ductal cells and outer myoepithelial cells. (C, D) HRAS wild-type sebaceous EMC. D, The staining is restricted to the inner sebaceous ductal cells, not the outer myoepithelial cells.

In contrast to the EMCs, most of the EMC-like salivary gland tumors were negative for RAS Q61R IHC (Fig. 4). However, one case each of PA and basal cell adenocarcinoma displayed diffuse and strong staining with coarsely granular cytoplasmic and a thick membranous pattern, which was slightly different from but hardly distinguishable from the “true-positive” immunoreactivity seen in HRAS Q61R-mutated EMCs. In contrast to EMC, these cases were genetically HRAS, KRAS, and NRAS wild-type. Another AdCC case showed focal luminal cell signaling and was scored as 1+.

FIGURE 4.

The histological and RAS Q61R immunohistochemical features of epithelial-myoepithelial carcinoma-like salivary gland tumors: adenoid cystic carcinoma (A, B) and pleomorphic adenoma (C, D). (A, C) Both tumors display a bi-layered tubular pattern with clear myoepithelial cells. (B, D) RAS Q61R immunohistochemistry is negative for both tumors.

Calculated from the whole cohort including 83 EMCs and 66 EMC-like salivary gland tumors, both the sensitivity and specificity of RAS Q61R IHC to the HRAS Q61R mutation analysis were 98%. Furthermore, the sensitivity of RAS Q61R IHC for the histologic diagnosis of EMC was 65%, while the specificity was 97% (Table 5).

TABLE 5.

Sensitivity and Specificity of RAS Q61R Immunohistochemistry

	RAS Q61R Immunohistochemistry		Total
	Positive	Negative
HRAS Q61R mutated EMC	54	1	55
Other EMCs + EMC mimics	2	92	94
Total	56	93	149
Sensitivity 98%, Specificity 98%

	RAS Q61R Immunohistochemistry		Total
	Positive	Negative
EMC	54	29	83
EMC mimics	2	64	66
Total	56	93	149

EMC indicates epithelial-myoepithelial carcinoma.

Sensitivity 65%, Specificity 97%

Discussion

In the last few decades, various tumor-specific genetic alterations that characterize particular histological entities have been discovered in salivary gland neoplasms.^{19, 20} There are mainly two types of such alterations: gene rearrangements creating fusion genes, such as PLAG1/HMGA2 rearrangement in PA, CRTC1/3-MAML2 in mucoepidermoid carcinoma, MYB/MYBL1-NFIB in AdCC, and ETV6-NTRK3/RET/MET in secretory carcinoma; and hotspot point mutations, such as PRKD1 E710D in polymorphous adenocarcinoma, CTNNB1 I35T in basal cell adenoma/adenocarcinoma, and BRAF V600E in sialadenoma papilliferum.^20–24 Another genetic change worth noting is HRAS hotspot mutations, which have recently been reported in EMC; due to their high specificity and sensitivity, these mutations have become a diagnostically important finding for this tumor.^{2, 7, 8} Although molecular testing has greatly enhanced the ability to obtain an accurate diagnosis, this approach is not always feasible compared to IHC in pathological laboratories. Thus, in this large cohort study, we used IHC to show that RAS Q61R is an excellent and useful marker for EMC that accurately reflects the HRAS Q61R mutation status.

IHC surrogates to detect proteins corresponding to genetic alterations for the pathological diagnosis of salivary gland tumors have been eagerly sought by institutions and hospitals where molecular analyses cannot be performed. Such surrogates include PLAG1/HMGA2 in PA, MYB in AdCC, pan-Trk in secretory carcinoma, and HER2 and p53 in salivary duct carcinoma.^25–30 However, the specificity of IHC for these markers is controversial.^{25, 31} Furthermore, mutant-specific IHC, e.g. that for BRAF V600E, has also been utilized.^{23, 24} The antibody used in the present study has been reported to have nearly 100% sensitivity and specificity for the RAS family protein Q61R mutant.^{15, 16, 18, 32–35} In fact, several malignant tumors harboring the RAS Q61R mutation, such as malignant melanoma (NRAS), colorectal carcinoma (KRAS), and medullary thyroid carcinoma (HRAS), demonstrate diffuse cytoplasmic and membranous immunoreactivity^{14, 16} corresponding to the localization of RAS protein in the tumor cells.³⁶ Our findings for RAS Q61R IHC in EMC closely matched those for the HRAS Q61R mutation status and the expression pattern described in other tumors.

EMC is a well-recognized low-grade malignant tumor of the salivary gland. The characteristic histology of classic EMC with a biphasic glandular formation with eosinophilic ductal luminal cells and clear abluminal myoepithelial cells may facilitate a correct diagnosis. However, since EMC frequently exhibits histologic variations, including cribriform, basaloid, sebaceous, oncocytic/apocrine, papillary-cystic, double-clear, squamous, psammomatous, and Verocay-like patterns, it can easily be confused with other salivary gland tumor entities.² For example, a cribriform pattern, which is the most common variations in EMC,² is a well-known hallmark of AdCC. EMC with a predominantly basaloid appearance may be misdiagnosed as basal cell adenoma/adenocarcinoma.³⁷ In addition, EMC can undergo high-grade transformation, which is associated with a worse prognosis.^{3, 38–41} In this study, diffuse and strong cytoplasmic and membranous RAS Q61R immunopositivity was observed in EMC, independently of specific histologic variations (Supplementary Table S4, Supplemental Digital Content 4). One point to be aware of regarding RAS Q61R IHC in histologic variants of EMC, however, is the strong staining seen in sebaceous luminal cells. Sebaceous differentiation is a relatively common histological variation of EMC detected in 13% to 17% of cases in previous studies as well as in the present cohort.^2–4 Since sebaceous differentiation may occur, as it is related to luminal ductal cells rather than myoepithelial cells, the true RAS Q61R IHC expression must be determined by focusing solely on the non-sebaceous myoepithelial cells (Fig. 4).⁴ Sebaceous change has usually been restricted to the limited foci in the tumor with few exceptions in previous reports.⁴ Furthermore, HRAS mutations as well as positive RAS Q61R IHC results were detected only in de novo EMC (71%), not in EMC ex PA, suggesting that the oncogenic pathway differs depending on the histologic origin of EMC.^{2, 7, 8} Therefore, RAS Q61R IHC might be of limited use for the diagnosis of EMC ex PA.

Many benign or malignant salivary gland tumors can exhibit clear myoepithelial features, which make the diagnosis of EMC challenging, particularly when faced with a small pathologic sample, such as that from a core biopsy or intraoperative consultation.⁴² Because IHC findings using markers of ductal epithelial and myoepithelial differentiation are similar among these tumors, the development of novel and reliable IHC markers that can lead to an accurate diagnosis of EMC are of great importance. Given the high specificity and reliability of RAS Q61R IHC in discriminating EMC from its mimics, RAS Q61R-positive staining strongly favors EMC. RAS Q61R IHC may be a useful diagnostic tool for EMC in cases with limited tissue samples (e.g. cell blocks for aspiration and core biopsies) where invasive carcinoma may not be evident and/or in tumors that are circumscribed and encapsulated without evidence of invasion; however, further studies are needed to verify the practical utility of this approach in such situations. Two EMC-like salivary gland tumor cases lacking an HRAS Q61R mutation, one case each of PA and basal cell adenocarcinoma, were diffusely positive for RAS Q61R IHC in the current study. In these cases, KRAS and NRAS as well as HRAS were wild-type, so the immunoreactivity was considered to be non-specific. Although we have noticed slight differences in the staining pattern from that of the true-positive cases, it must be kept in mind that there are occasional pseudo-positive cases that cannot be discriminated by IHC alone.

HRAS Q61R mutations have also been reported in some salivary gland tumor entities other than EMC, including salivary duct carcinoma (12/140 cases, 8.6%),^{30, 43} sialadenoma papilliferum (2/10 cases, 20%), and intraductal papillary mucinous neoplasm (4/9 cases, 44%).²³ Although these tumors are histomorphologically distinct and differ from EMC, careful exclusion is required when the tumor is positive for RAS Q61R IHC.

The interpretation of RAS Q61R IHC was basically straightforward; the diffuse cytoplasmic and membranous expression was confined to the myoepithelial cells, and the luminal cells were usually negative. Adenomyoepithelioma of the breast shares a similar histomorphology and genetic profile to EMC.^{17, 44, 45} One study examined the correlation between the HRAS mutation status and the RAS Q61R IHC findings in mammary adenomyoepithelioma and found that 5 of 7 cases harboring an HRAS Q61R mutation were immunoreactive for RAS Q61R IHC, whereas 2 cases with an HRAS Q61R mutation and 17 HRAS wild-type adenomyoepithelioma cases showed negative immunostaining.¹⁷ The immunoreactivity was almost always restricted to the myoepithelial cells, which was in line with our findings for EMC.¹⁷ Furthermore, the investigators microdissected epithelial cells and myoepithelial cells separately and identified HRAS Q61R mutations in both type of cells.¹⁷ The discrepancy in the expression by the nature of the tumor cells might depend on the difference in the mutant HRAS protein level resulting from posttranscriptional regulation.

The RAS mutation status as well as RAS mutant-specific protein expression might have future therapeutic relevance. While most EMCs are of low-grade malignancy and can be treated with surgical resection alone,^{46, 47} for EMC patients with high-grade transformation or metastatic/recurrent tumor, RAS-targeting systemic therapy may be of benefit. Although the development of RAS mutation-specific covalent inhibitor is restricted to the KRAS G12C mutation, RAS-effector interaction inhibitors, including MEK inhibitor binimetinib, might be effective against HRAS-mutated tumors.^{13, 48, 49}

In conclusion, RAS Q61R IHC is an accurate and cost-effective method for detecting an HRAS Q61R mutation in EMC. Diffuse positive cytoplasmic and membranous staining strongly suggests the diagnosis of EMC when the histology is also compatible with this tumor entity. RAS Q61R IHC has the potential to contribute to the correct diagnosis of EMC in challenging practical pathology settings.

Supplementary Material

Supplemental Table S1.pdf

Supplemental Digital Content 1. Supplementary Table 1: PCR Primers Used for Sanger Sequencing. pdf

Supplemental Table S2.pdf

Supplemental Digital Content 2. Supplementary Table 2: Antibodies Used for Immunohistochemistry. pdf

Supplemental Table S3.pdf

Supplemental Digital Content 3. Supplementary Table 3: Clinical Characteristics of Salivary Gland Tumors with Partial EMC-Like Component. pdf

Supplemental Table S4.pdf

Supplemental Digital Content 4. Supplementary Table 4: Results of RAS Q61R Immunohistochemistry in Epithelial-Myoepithelial Carcinoma Variants. pdf