NR4A3 Immunohistochemistry Reliably Discriminates Acinic Cell Carcinoma from Mimics

Abstract

Acinic cell carcinoma (AciCC) harbors a recurrent t(4;9)(q13;q31) translocation, which leads to upregulation of Nuclear Receptor Subfamily 4 Group A Member 3 (NR4A3). Previous work on tissue microarrays suggests that NR4A3 immunohistochemistry (IHC) may be useful in the diagnosis of AciCC. Thus far, only a single study has evaluated the utility of NR4A3 immunohistochemistry (IHC) in the diagnosis of AciCC, using a tissue microarray to assess most non-AciCC tumor types. Herein we evaluate the diagnostic performance of NR4A3 IHC for AciCC in a large cohort of 157 salivary gland tumors, using whole tissue sections. The cohort consisted of 37 AciCC (6 of them (16%) with high grade transformation), 30 secretory carcinomas (SC), and 90 additional salivary gland tumors, including mucoepidermoid carcinomas (MEC), polymorphous adenocarcinomas (PAC), pleomorphic adenomas (PA), salivary duct carcinomas (SDC), and adenoid cystic carcinomas (AdCC). NR4A3 nuclear staining by IHC was considered positive if present in more than 5% of tumor cells. Overall, 92% of AciCC (34/37) expressed NR4A3 by IHC, with strong (89%) or moderate (3%) nuclear staining, yielding a sensitivity of 92%. IHC detected NR4A3 expression in all cases of recurrent/metastatic AciCC and tumors with high grade transformation. Importantly, all SC were negative for NR4A3 IHC, with no staining in 28/30 cases and weak focal staining, in < 5% of cells, in 2/30 (7%). Similarly, all MEC (20/20), SDC (20/20) and AdCC (10/10) were negative for NR4A3 by IHC, as were most PA (18/20; 15%) and PAC (18/20; 5%). Two PA and two PAC showed multifocal expression of NR4A3 in more than 5% of cells, of weak intensity in 3 cases and moderate in 1 PAC, yielding an overall specificity of 97% for NR4A3 IHC for the diagnosis of AciCC. In conclusion, NR4A3 is a highly sensitive and specific immunohistochemical marker for AciCC; moderate to strong and/or diffuse NR4A3 expression is a consistent and diagnostic feature of AciCC.

Introduction

Acinic cell carcinoma (AciCC) is a malignant salivary neoplasm which demonstrates serous acinar differentiation. The histologic features of these tumors are often characteristic, with serous acinar cells containing basophilic zymogen granules admixed with intercalated duct-type cells with variably clear, vacuolated, or oncocytoid cytoplasm. Solid, microcystic, follicular, or papillary architecture typically predominates [1]. Although a morphologic diagnosis is often straightforward, AciCC can also demonstrate significant histologic overlap with other low to intermediate grade salivary tumors, particularly when serous acinar differentiation is subtle. Secretory carcinoma (SC), which demonstrates a recurrent ETV6–NTRK3 fusion [2], is perhaps the most notorious histologic mimic of AciCC. Following the initial histologic and molecular description of SC, retrospective reviews of AciCC demonstrated a high proportion of SC among non-parotid and/or zymogen granule-poor cases [2, 3, 4, 5].

The histochemical and immumohistochemical markers traditionally used to aid in the diagnosis of AciCC are non-specific. The characteristic zymogen granules are positive for Periodic acid–Schiff (PAS) and diastase-resistant, however the extracellular secretory material of SC is also PAS-positive, diastase-resistant. Immunohistochemistry (IHC) for SOX10 [6] and DOG1 [7] are frequently positive in AciCC (the latter in a characteristic but somewhat subjective canalicular pattern), although both markers have limited discriminatory value and are positive in a range of salivary tumors. AciCC is negative for p63 and S-100. While SC and other salivary neoplasms can be excluded by identification of their defining immunohistochemical or molecular profiles, this often requires a significant number of ancillary studies, some of which are unavailable or cost prohibitive at most institutions. Recently, a t(4;9)(q13;q31) recurrent chromosomal rearrangement was found to occur at high frequency in AciCC [8]. As a result of this rearrangement, Nuclear Receptor Subfamily 4 Group A Member 3 (NR4A3) becomes constitutively upregulated through enhancer hijacking. In the initial description of this rearrangement and a subsequent study focusing on use of NR4A3 IHC as a diagnostic tool in AciCC, Haller et al. demonstrated NR4A3 IHC-positivity in > 97% of AciCC [8, 9]. In this latter study, NR4A3 IHC was performed on AciCC, SC, and a tissue microarray containing other salivary tumors. The goal of this study was to further evaluate the performance of NR4A3 IHC in AciCC and to broaden the scope of other salivary tumors evaluated using whole tissue sections.

Methods

Study Population and Data Acquisition

Approval was obtained from the Brigham and Women's Hospital Institutional Review Board. Cases of AciCC and secretory carcinoma (SC) with material available for additional studies were retrieved from the surgical pathology files (n = 37 and 30, respectively). Additionally, 20 mucoepidermoid carcinomas (MEC), 20 polymorphous adenocarcinomas (PAC), 20 pleomorphic adenomas (PA), 20 salivary duct carcinomas (SDC), and 10 adenoid cystic carcinomas (AdCC) were obtained. Clinicopathologic information, including patient demographics, primary tumor location, tumor site sampled (primary, recurrence, metastasis), pathologic diagnosis, histochemical and IHC results, and presence of high grade transformation was obtained from the patient medical record and surgical pathology reports.

Immunohistochemistry (IHC) for NR4A3

NR4A3 IHC was performed on 4-µm-thick formalin-fixed paraffin-embedded (FFPE) whole-tissue sections using pressure cooker antigen retrieval in pH 6.1 citrate buffer (Target Retrieval Solution, Dako, Carpinteria, CA). The NOR1/NR4A3 mouse monoclonal antibody (clone H-7, 1:50 dilution, Santa Cruz Biotechnology, Dallas, TX, USA) was applied followed by the Novolink Polymer Detection System (Leica, Buffalo Grove, IL, USA). Nuclear staining was semi-quantitatively scored by intensity (weak, moderate, strong) and extent (0: 0%, 1 + : < 5%, 2 + : 5–50%, and 3 + : > 50%). At least 2 + staining extent and/or intensity in tumor cells was considered a positive result.

Fluorescence In Situ Hybridization (FISH) for NR4A3

Interphase FISH analysis was performed according to standard protocols on 4-µm-thick FFPE tissue sections from 5 cases of AciCC (including 2 cases showing high grade transformation) which stained positively for NR4A3. The presence of NR4A3 rearrangement was evaluated using a commercially available dual-color break-apart probe set (ZytoLight SPEC NR4A3 Dual Color Break Apart Probe, Z-2145–50, ZytoVision GmbH, Bremerhaven, Germany) as described in the literature [8]. Signals separated by > 3 signal diameters were considered split signals. The slides were reviewed manually by a pathologist (A.M.E.) and at least 50 tumor nuclei were evaluated for each case. One case was technically unsuccessful. Tumors with more than 30% nuclei showing isolated signals or split signals were considered positive for NR4A3 rearrangement.

Results

Clinicopathologic Characteristics

Overall, 157 tumor samples from 153 patients were obtained, including 37 cases of AciCC from 35 patients. 30 (81%) AciCC cases were from sampling of the primary tumor, and 7 (19%) were recurrent or metastatic tumors. The primary tumor site from the 35 patients was the parotid gland in 31 (89%) cases and minor salivary gland in 4 (11%) cases. Of the 37 tissue samples, 31(84%) AciCC were histologically low grade, while 6 (16%) demonstrated high grade transformation. Additionally, 30 cases of SC were identified from 29 patients. Most tumors originated in the parotid (76%), while a smaller subset (24%) arose from other sites, including the submandibular gland (7%). Of SC cases, 26 (87%) represented primary tumors and 4 (13%) cases were recurrent or metastatic disease. All cases were histologically low grade. The clinicopathologic characteristics of the study cohort, as well as those of MEC, PAC, PA, SDC, and AdCC, are summarized in Table 1.

Table 1.

Clinicopathologic characteristics

	Acinic Cell carcinoma (n = 37)*	Secretory carcinoma (n = 30)*	Mucoepidermoid carcinoma (n = 20)	Polymorphous adenocarcinoma (n = 20)*	Pleomorphic adenoma (n = 20)	Salivary duct carcinoma (n = 20)	Adenoid cystic carcinoma (n = 10)
Sex [n (%)]
Female	22 (63)	11 (38)	13 (65)	10 (53)	14 (70)	9 (45)	6 (60)
Male	13 (37)	18 (62)	7 (35)	9 (47)	6 (30)	11 (55)	4 (40)
Age (years)
Mean	53	51	51	53	51	69	53
Range	12–86	19–83	17–75	26–83	30–74	54–87	28–90
Primary tumor site [n (%)]
Parotid	31 (89)	22 (76)	7 (35)	0 (0)	18 (90)	12 (60)	3 (30)
Submandibular	0 (0)	2 (7)	0 (0)	0 (0)	2 (10)	6 (30)	2 (20)
Sublingual	0 (0)	0 (0)	1 (5)	0 (0)	0 (0)	0 (0)	0 (0)
Other	4 (11)	5 (17)	12 (60)	19 (100)	0 (0)	2 (10)	5 (50)
Tumor sampled [n (%)]
Primary	30 (81)	26 (87)	20 (100)	20 (100)	20 (100)	16 (80)	20 (100)
Recurrence/metastasis	7 (19)	4 (13)	0 (0)	0 (0)	0 (0)	4 (20)	0 (0)
High grade [n (%)]	6 (16)	0 (0)	1 (5)	0 (0)	N/A	N/A	0 (0)

^*N refers to the number of tumor samples examined. One patient had at least two different tumor samples in each of the three cohorts indicated. The sex, age, and primary tumor site statistics are based on the number of patients, whereas the statistics for tumor sampled and tumor grade refer to the number of tumor samples

NR4A3 Immunohistochemistry in Acinic Cell Carcinoma and Other Tumors

Overall, immunohistochemistry (IHC) for NR4A3 was positive in 34 (92%) AciCC. Of these cases, there was strong 3 + staining in 30 (81%) cases, moderate 3 + staining in 1 (3%) case, and strong 2 + staining in 3 (8%) cases (Fig. 1a–f). All cases of recurrent or metastatic AciCC and tumors with high grade transformation were positive for NR4A3 (Fig. 1d–f). Of the 3 IHC-negative cases (Fig. 1g–i), 1 was from the parotid gland, while 2 were from minor salivary gland. All cases demonstrated either serous acinar differentiation morphologically or had supporting histochemical and IHC studies to support the diagnosis (including the presence of PAS-positive, diastase-resistant granules in 13 cases and DOG1 canalicular staining in 15 cases). Other studies performed in the AciCC cohort are summarized in Table 2.

Fig. 1.

Examples of acinic cell carcinoma (a–c) and acinic cell carcinoma with high grade transformation (d–f) with positive NR4A3 immunohistochemistry. NR4A3 was negative in few cases (g–i)

Table 2.

Additional Ancillary Studies in Acinic Cell Carcinoma Cohort

	Positive	Negative	Total
PAS-D (n)	13	0	13
DOG-1 (n)	15	0	15
p63/p40 (n)	1a	5	6
S100 (n)	2b	10	12
Mammaglobin (n)	0	8	8
Pan-Trk (n)	0	3	3
PLAG1 (n)	0	3	3

^aOne case demonstrated focal p63 positivity

^bOne case demonstrated focal S100 positivity. Another case with high grade transformation demonstrated more extensive staining but had a low grade component morphologically consistent with acinic cell carcinoma

Of the 130 non-AciCC cases, 4 (3%) cases were positive for NR4A3 by IHC, including 3 of 20 (15%) PAC and 1 of 20 (5%) PA (see Table 3). However, none demonstrated more than 2 + staining. Three of the cases had weak staining (Fig. 2a, b), while 1 case of PAC had moderate 2 + staining (Fig. 2c, d). All cases of SC were negative (Fig. 2e, f), with 28 (93%) cases demonstrating absence of staining and 2 (7%) cases showing weak 1 + staining. Similarly, all cases of MEC, SDC, and AdCC were negative for NR4A3.

Table 3.

NR4A3 Immunohistochemistry in Acinic Cell Carcinoma and Other Salivary Tumors

	Acinic Cell carcinoma (n = 37)	Secretory carcinoma (n = 30)	Mucoepidermoid carcinoma (n = 20)	Polymorphous adenocarcinoma (n = 20)	Pleomorphic adenoma (n = 20)	Salivary duct carcinoma (n = 20)	Adenoid cystic carcinoma (n = 10)
Staining extent and intensity [n (%)]
0	3 (8)	28 (93)	16 (80)	10 (50)	19 (95)	18 (90)	10 (100)
1+ , total	0 (0)	2 (7)	4 (20)	7 (35)	0 (0)	2 (10)	0 (0)
Weak/mod/strong	–	2/0/0	4/0/0	6/1/0	–	2/0/0	–
2+ , total	3 (8)	0 (0)	0 (0)	3 (15)	1 (5)	0 (0)	0 (0)
Weak/mod/strong	0/0/3	–	–	2/1/0	1/0/0	–	–
3+ , total	31 (84)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
Weak/mod/strong	0/1/30	–	–	–	–	–	–
Overall IHC result [n (%)]
Positive (≥ 2 +)	34 (92)	0 (0)	0 (0)	3 (15)	1 (5)	0 (0)	0 (0)
Negative (< 2 +)	3 (8)	30 (100)	20 (100)	17 (85)	19 (95)	20 (100)	10 (100)

0: 0%; 1 + : < 5%; 2 + : 5–50%; 3 + : > 50%; Mod: moderate; IHC: immunohistochemistry

Fig. 2.

Examples of NR4A3 immunohistochemistry in other salivary tumors. Rarely ≥ 5% staining was seen, such as in this pleomorphic adenoma with weak staining (a, b) and this polymorphous adenocarcinoma with moderate staining (c, d). All secretory carcinomas were negative when using a 5% cutoff (e, f)

Fluorescence In Situ Hybridization for NR4A3

Four AciCCs that were positive for NR4A3 IHC were successfully evaluated by FISH. NR4A3 rearrangements were observed in 3/4 cases (Fig. 3), including 1 tumor with high grade transformation and 1 distant metastasis. One tumor showed apparently normal fusion signals. All low grade tumors (including 2 with demonstrable NR4A3 rearrangement and 1 without) were diagnosed based on morphologic features. The case with high grade transformation showed conventional areas, was positive for DOG-1 and negative for S100 and mammaglobin.

Fig. 3.

Dual-color break-apart interphase NR4A3 FISH in AciCC. Split red and green hybridization signals suggest the presence of a genomic NR4A3 rearrangement in this tumor. Note that in addition to a separate green signal, a small fragment of the green probe co-localizes with a larger red probe in many cells, suggesting that the genomic breakpoint in this tumor is located within the target sequence of the green probe

Discussion

In 2019, acinic cell carcinoma (AciCC) was found to harbor the recurrent translocation t(4;9)(q13;q31) which leads to the upregulation of the NR4A3 transcription factor. The clinical utility of NR4A3 IHC in the diagnosis of AciCC [9] has been suggested by a study that showed high sensitivity and specificity for AciCC, particularly in the discrimination from SC; however, evaluation of other salivary tumor types was restricted to a tissue microarray that contained a predominance of adenoid cystic carcinomas and relatively fewer other tumors in the differential diagnosis. In this study, we comprehensively evaluated the performance of NR4A3 IHC in AciCC and a large number of other salivary tumors using whole tissue sections.

In concordance with the prior studies by Haller et al. [8, 9], we observed that NR4A3 IHC was positive in the vast majority of AciCC (92%), with most cases demonstrating strong, diffuse (3 + or > 50%) nuclear staining. This included all cases of AciCC which were recurrent/metastatic and those that demonstrated high grade transformation. While the majority of AciCC are low grade tumors, high grade transformation may infrequently occur. These high grade tumors can demonstrate aggressive behavior, frequently with destructive local growth and metastases [3, 10,20]. Diagnosis may be difficult unless a more recognizable conventional component is also present, and the NR4A3 IHC may therefore be useful to confirm the diagnosis and to distinguish from other high grade salivary tumors, such as salivary duct carcinoma.

Of the three cases of AciCC which were negative for NR4A3 IHC, 1 occurred in the parotid and morphologically demonstrated serous acinar differentiation (Fig. 1g–i). This case was also negative for S100, mammaglobin, and pan-Trk IHC. The other 2 cases arose in minor salivary gland; although they did not have classic morphologic features of AciCC, both cases had PAS-positive, diastase-resistant granules and were negative for mammaglobin and pan-Trk IHC. S100 was negative in 1 case and only had focal positivity in the other case. P63 was also performed on 1 of the 2 cases and was negative. While AciCC typically occurs in the parotid, it may also arise less commonly in other major and minor salivary glands [21, 23]. In the initial studies by Haller et al., only tumors arising from the parotid were evaluated [8, 9]. In our study, NR4A3 staining was seen in 30/31 (97%) cases of AciCC from the parotid and 2/4 (50%) from minor salivary sites.

Interestingly, detection of NR4A3 expression by IHC did not perfectly correlate with the presence of NR4A3 rearrangement, as assessed by FISH; NR4A3 rearrangements were detected in 3/4 IHC-positive cases. This is not surprising, given that enhancer hijacking is the mechanism of NR4A3 upregulation in AciCC and the location of the 9q31 genomic breakpoints in relation to these FISH probes, as discussed by Haller et al. [9]. In AciCC, certain NR4A3 rearrangements may be missed by FISH and a negative result in this context should be interpreted carefully. In addition, a small subset of AciCC (~ 3–15%) harbor alternative HTN3-MSANTD3 oncogenic fusions [9, 24, 25], which lead to comparable NR4A3 upregulation but are not detectable by NR4A3 FISH [9, 25]. Given that NR4A3 overexpression represents a critical biological feature of AciCC, regardless the underlying genomic mechanism, NR4A3 IHC is the most sensitive and cost-effective diagnostic tool available in the clinical setting.

Staining for NR4A3 was rare in other salivary tumors, with positivity (at least 2 + , or > 5% staining) seen in only few cases of PA (1 of 20) and PAC (3 of 20). Three of these cases demonstrated only weak 2 + staining, while one case of PAC demonstrated moderate 2 + staining (Fig. 2c, d). Overall, when using a cutoff of 5%, the sensitivity and specificity of NR4A3 for AciCC was 92% and 97%, respectively. Notably, all cases of secretory carcinoma (SC) were negative for NR4AA3 IHC. Prior to the recognition of SC as a separate entity in the salivary gland, many cases had been classified as zymogen-poor AciCC [2–5]. SC, which are defined by ETV6-NTRK3 fusions in most cases, typically show immunoreactivity for S100 and mammaglobin [2, 26]. Co-expression of these markers in tumors in conjunction with characteristic morphology has been found to correlate strongly with the presence of ETV6 rearrangement in SC [26], although S-100 and mammaglobin may be occasionally positive in other salivary tumors as well [27–29]. More recently, pan-TRK IHC has also been used to distinguish SC from AciCC and other mimic, although the reported sensitivities and specificities have been variable [30–32]. DOG1 IHC may be used in the differential diagnosis, which is typically negative in SC and positive in AciCC in an apical membranous pattern; however DOG1 staining can be show non-specific staining in a range of other salivary tumors including SC [7]. The availability of NR4A3 IHC may therefore be helpful for definitive diagnosis of AciCC, particularly in small biopsies and in challenging cases where morphology or other IHC results are not straightforward.

Additionally, all cases of salivary duct carcinoma (SDC), adenoid cystic carcinoma (AdCC), and mucoepidermoid carcinoma (MEC) tested in this study were also negative for NR4A3 IHC. Although SDC and AdCC are less often mistaken for AciCC based on morphology and clinical behavior, ancillary studies may be necessary, especially in small biopsies or in tumors with high grade transformation where the conventional precursor tumor is not well represented. MEC, which is characterized by recurrent CRCT1-MAML2 fusions [33,34] may mimic AciCC, particularly when the characteristic mucinous component is not readily identified. In cases where molecular testing is unavailable or not possible due to limited material, NR4A3 IHC may be useful to exclude a diagnosis of AciCC.

In conclusion, we have shown that nuclear staining for NR4A3 IHC demonstrates high sensitivity and specificity for AciCC. Focal or weak staining may be seen in rare non-AciCC cases, thus interpretation of IHC is recommended in the context of morphologic examination. Molecular testing is often costly, and is less sensitive than IHC given that NR4A3 is upregulated in AciCC regardless of underlying genomic alterations. NR4A3 IHC can serve as a useful adjunct in the diagnosis of AciCC and can help to distinguish AciCC from SC and other primary salivary tumors.

Funding

This work was funded by departmental resources.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflicts of interest.