If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Tissue Pathology and Diagnostic Oncology, NSW Health Pathology, Royal Prince Alfred Hospital, Sydney, NSW, AustraliaSydney Medical School, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, Australia
Department of Tissue Pathology and Diagnostic Oncology, NSW Health Pathology, Royal Prince Alfred Hospital, Sydney, NSW, AustraliaSydney Medical School, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, AustraliaDiscipline of Pathology, School of Medicine, Western Sydney University, Campbelltown, NSW, AustraliaDepartment of Anatomical Pathology, Liverpool Hospital, Liverpool, NSW, AustraliaCancer Pathology Laboratory, Ingham Institute for Applied Medical Research, Liverpool, NSW, AustraliaCONCERT Biobank, Ingham Institute for Applied Medical Research, Liverpool, NSW, AustraliaSouth Western Sydney Clinical School, University of New South Wales, Liverpool, NSW, Australia
Department of Head and Neck Surgery, Chris O'Brien Lifehouse, Sydney, NSW, AustraliaDepartment of Otolaryngology – Head and Neck Surgery, Faculty of Medicine and Health Sciences, Macquarie University, Macquarie Park, NSW, Australia
Department of Tissue Pathology and Diagnostic Oncology, NSW Health Pathology, Royal Prince Alfred Hospital, Sydney, NSW, AustraliaSydney Medical School, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, Australia
Department of Head and Neck Surgery, Chris O'Brien Lifehouse, Sydney, NSW, AustraliaSydney Medical School, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, AustraliaDepartment of Otolaryngology – Head and Neck Surgery, Faculty of Medicine and Health Sciences, Macquarie University, Macquarie Park, NSW, Australia
Department of Head and Neck Surgery, Chris O'Brien Lifehouse, Sydney, NSW, AustraliaSydney Medical School, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, Australia
Department of Head and Neck Surgery, Chris O'Brien Lifehouse, Sydney, NSW, AustraliaSydney Medical School, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, AustraliaRoyal Prince Alfred Institute of Academic Surgery, Sydney Local Health District, Sydney, NSW, Australia
Address for correspondence: Prof Ruta Gupta, Department of Tissue Pathology and Diagnostic Oncology, NSW Health Pathology, Building 94, Royal Prince Alfred Hospital, John Hopkins Drive, Camperdown, NSW 2050. Australia.
Department of Tissue Pathology and Diagnostic Oncology, NSW Health Pathology, Royal Prince Alfred Hospital, Sydney, NSW, AustraliaSydney Medical School, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, Australia
Adenoid cystic carcinoma (ACC) is one of the most common primary salivary gland cancers. ACC has several benign and malignant mimics amongst salivary gland neoplasms. An accurate diagnosis of ACC is essential for optimal management of the patients and their follow-up. Upregulation of MYB has been described in 85–90% of ACC, but not in other salivary gland neoplasms. In ACC, MYB upregulation can occur as a result of a genetic rearrangement t(6;9) (q22–23;p23–24), MYB copy number variation (CNV), or enhancer hijacking of MYB. All mechanisms of MYB upregulation result in increased RNA transcription that can be detected using RNA in situ hybridisation (ISH) methods.
In this study, utilising 138 primary salivary gland neoplasms including 78 ACC, we evaluate the diagnostic utility of MYB RNA ISH for distinguishing ACC from other primary salivary gland neoplasms with a prominent cribriform architecture including pleomorphic adenoma, basal cell adenoma, basal cell adenocarcinoma, epithelial myoepithelial carcinoma, and polymorphous adenocarcinoma. Fluorescent in situ hybridisation and next generation sequencing were also performed to evaluate the sensitivity and specificity of RNA ISH for detecting increased MYB RNA when MYB gene alterations were present. Detection of MYB RNA has 92.3% sensitivity and 98.2% specificity for a diagnosis of ACC amongst salivary gland neoplasms. The sensitivity of MYB RNA detection by ISH (92.3%) is significantly higher than that of the FISH MYB break-apart probe (42%) for ACC. Next generation sequencing did not demonstrate MYB alterations in cases that lacked MYB RNA overexpression indicating high sensitivity of MYB RNA ISH for detecting MYB gene alterations. The possibility that the sensitivity may be higher in clinical practice with contemporary samples as compared with older retrospective tissue samples with RNA degradation is not entirely excluded. In addition to the high sensitivity and specificity, MYB RNA testing can be performed using standard IHC platforms and protocols and evaluated using brightfield microscopy making it a time and cost-efficient diagnostic tool in routine clinical practice.
ACC is a biphasic salivary gland carcinoma that exhibits overlapping morphological and immunophenotypic characteristics with many salivary gland neoplasms including benign tumours like pleomorphic adenoma (PA) and basal cell adenoma (BCA), and malignant tumours such as epithelial myoepithelial carcinoma (EMC), polymorphous adenocarcinoma (PAC), and basal cell adenocarcinoma (BCAC).
Unlike its mimics, ACC has a high propensity for infiltrative growth and perineural invasion, thus requiring functionally morbid surgical resection followed by radiotherapy for local control.
Therefore, an accurate distinction of ACC from its less aggressive mimics is critical at the time of initial diagnosis.
Upregulation of MYB, a nuclear transcription factor responsible for cell proliferation and survival, has been observed in 85–90% of ACC, but not in other salivary gland neoplasms.
MYB upregulation can occur as a result of a genetic rearrangement t(6;9)(q22–23;p23–24) leading to MYB-NFIB fusion in 40–60% ACC, MYB copy number variation (CNV) in 40–50% tumours and enhancer hijacking of MYB in a minority of ACC.
A variety of methods including immunohistochemistry (IHC), fluorescent in situ hybridisation (FISH) and next generation sequencing (NGS) have been used to detect MYB alterations in ACC with variable sensitivity and specificity. MYB IHC has been described to have a sensitivity of 82–85% for ACC but a low specificity of 54%.
for ACC, but a sensitivity of 40–60% as it can only detect MYB gene rearrangement. Detection of MYB through NGS is expensive, may not be universally available, and requires evaluation of both RNA and DNA to detect MYB fusion and CNV. However, MYB upregulation leads to increased MYB RNA production (Fig 1), irrespective of the underlying genetic mechanism. Rooper et al. have demonstrated that MYB RNA detection had 92% sensitivity and 95% specificity in distinguishing ACC from other salivary tumours.
MYB RNA can be detected using in situ hybridisation (ISH) staining methods in a cost-efficient manner. Herein, we evaluate the diagnostic utility of MYB RNA detection by ISH in distinguishing ACC from other primary salivary gland neoplasms with a prominent cribriform architecture including a subset of PA, BCA, BCAC, EMC, and PAC. FISH and NGS were also performed to evaluate the sensitivity and specificity of RNA ISH for detecting increased MYB RNA. The current cohort represents international validation of the findings by Rooper et al.
and facilitates addition of MYB RNA detection by ISH into routine clinical diagnostic practice.
Materials and methods
Tissue samples and clinicopathological parameters
Patients with primary salivary gland neoplasms were identified through the prospectively maintained Sydney Head and Neck Cancer Institute (SHNCI) database following approval by the Human Research Ethics Committee. All tumours were classified according to the histological classification of salivary gland tumours as recommended in the 4th edition of the World Health Organization classification of head and neck tumours.
El-Naggar A.K. Chan J.K.C. Grandis J.R. Takata T. Sootweg P. World Health Organization Classification of Head and Neck Tumours. 4th ed. IARC Press,
Lyon2017e202
All clinicopathological and follow-up information was obtained from the database.
The haematoxylin and eosin (H&E) stained slides were reviewed to identify 138 cases of primary salivary gland neoplasms including ACC, biphasic tumours, tumours with prominent cribriform architecture or common mimics of ACC including PA, BCA, EMC, PAC, myoepithelial carcinoma, and BCAC. Examples of metastatic ACC (n=14) were also included to understand the utility of MYB RNA ISH staining in the diagnoses of metastatic ACC.
Following a review of H&E slides, areas with high tumour cellularity and low basement membrane matrix or necrosis were utilised for tissue microarray (TMA) construction. TMAs were constructed using Beecher Manual Tissue Microarrayer (Model MTA-1; Diagnostic Technology, Australia). Two cores of 1.0 mm diameter for each case were extracted from formalin-fixed, paraffin-embedded (FFPE), tissue blocks of the salivary gland tumour samples.
MYB RNA in situ hybridisation
MYB RNA ISH was performed on all cases using TMA and on a subset (20%) of whole sections to identify heterogeneity if any. The commercially available MYB RNA probe targeting bases 204–1289 from the human MYB gene (Cat. no. 419708; Advanced Cell Diagnostics, USA) was used. These bases are proximal to the common MYB fusion breakpoints and thus should be conserved regardless of rearrangement.
Hybridisation was performed on 3 μM sections using a Leica Bond III automated staining platform (Leica Biosystems, USA) according to a standardised ISH protocol provided by the manufacturer. A case of ACC from the submandibular gland demonstrating MYB rearrangement by FISH and MYB RNA expression was used as external control.
MYB RNA ISH overexpression was defined as per the criteria suggested by Rooper et al. and the presence of three or more punctate signals or one clumped aggregate of signals in the nucleus or cytoplasm of at least 30% of tumour cells.
MYB FISH was performed at our institution prior to the availability of MYB RNA ISH. Thus, all cases with MYB FISH were retrieved from our database (2015–2020) (n=33) to compare the clinical utility of MYB FISH and MYB RNA ISH.
MYB FISH was performed using a commercially available MYB dual colour break apart probe (ZTV-Z-2143-200; ZytoVision, Germany) for the gene locus 6q23.3.
Interphase nuclei with two fusion signals of one orange and one green fluorochrome were evaluated as normal. A signal pattern consisting of one orange and green fusion signal and one orange and one green signal separated from each other by at least a distance equal to the diameter of one signal was interpreted as representing MYB rearrangement. The signals were evaluated in 100 tumour nuclei and tumours showing rearrangement in >15% nuclei were reported as positive for MYB rearrangement.
Tumours that were morphologically diagnosed as ACC but showed a negative MYB RNA ISH were subjected to NGS. FFPE tissue samples from eight cases with at least 30% tumour cellularity were analysed with the Ion Torrent Oncomine Comprehensive Assay v3 (ThermoFisher Scientific, Australia), which covers 161 cancer genes including MYB and MYBL1 gene fusions. DNA and RNA extractions were performed with QIAamp AllPrep DNA/RNA FFPE Tissue Kit (Cat. no./ID 80234; Qiagen, Germany), and subsequent quantification was performed using Qubit dsDNA HS (High Sensitivity) Assay Kit for DNA and Qubit RNA HS Assay Kit for RNA (ThermoFisher Scientific). NGS was then performed with the Ion Torrent Genexus Integrated Sequencer using the Ion Torrent GX5 Chip (ThermoFisher Scientific).
Sequencing was performed to a mean coverage of >1000× for DNA targets with DNA on-target reads at >85% and uniformity of base coverage at >90%. The limit of detection was 5% variant allele frequency at coverage of 400×. For RNA sequencing, mapped reads of >1,000,000 with a mean read length of >90 were considered suitable for fusion analyses.
Bioinformatics involved the use of Ion Reporter software to analyse, filer, and annotate variants, which were reported according to HGVS nomenclature (www.varnomen.hgvs.org, version 20.05).
Statistical analysis
Statistical analysis was undertaken to determine the sensitivity and specificity of MYB RNA ISH in the diagnosis of ACC using Excel (MicroSoft Corporation, USA) and 2×2 tables.
Results
The study included a total of 138 cases of primary salivary gland neoplasms with prominent cribriform architecture. This included cases diagnosed as ACC (n=80), PA (n=13), BCA (n=15), EMC (n=14), PAC (n=8), myoepithelial carcinoma (n=2), BCAC (n=6) from all sites of the head and neck with a median follow up of 6.7 years (8 months–16.5 years).
Of the 80 cases diagnosed as ACC, 48 were women and 32 were men with a median age range of 58 years (19–85 years). Material from distant metastases was available in 14 patients. The clinicopathological data are provided in Table 1.
Table 1Clinicopathological characteristics of patients with adenoid cystic carcinoma
A total of 72 (90%) histologically diagnosed ACC cases showed MYB RNA overexpression, and eight (10%) histologically diagnosed as ACC cases did not demonstrate MYB RNA overexpression. The MYB RNA overexpression was concordant between cores from the same patient in the TMA. Similarly, there was concordance in MYB RNA overexpression between the cores and whole sections. The whole sections of eight cases that did not demonstrate MYB RNA overexpression on TMA were also negative.
Amongst the 72 patients with detectable MYB RNA overexpression, most demonstrated classic biphasic cribriform architecture (Fig. 2A) with cells showing scanty cytoplasm and angulated hyperchromatic nuclei, infiltrative growth pattern, and perineural invasion (PNI). In most instances, the MYB RNA signals were easily visualised at a magnification of 10× or 20× and accounted for >4 signals/cell or clumps (Fig. 2B). Only one case with typical morphological appearance (Fig. 2C) showed focal and pale signals requiring evaluation at 40× magnification (Fig. 2D).
Fig. 2(A) Classic morphological appearance of adenoid cystic carcinoma with cribriform architecture and biphasic histology (H&E, 10×). (B) Easily identifiable MYB RNA signals as four or more brown signals and clusters. This was the most frequently observed pattern readily evaluated at 10× magnification. (C) Classic morphologic appearance of adenoid cystic carcinoma with perineural involvement (H&E, 10×). (D) Cells with >3 pale signals in >30% of the tumour cells. This pattern was rarely seen in this cohort and required scanning at 40× (arrows indicate).
MYB RNA expression was also observed in ACC with unconventional morphological appearance (Fig. 3). Small biopsies from sites with minor salivary gland tissue and metastatic deposits in the liver or bone marrow were over-represented amongst these cases. A morphological spectrum including tubules, solid or angulated nests (Fig. 3A,B), cytoplasmic clearing (Fig. 3C,D), signet ring morphology (Fig. 3E,F) and high grade transformation were seen. At the time of initial diagnosis, these cases had been histologically favoured to represent ACC in light of the clinical history or due to the presence of focal cribriform architecture or a biphasic immunohistochemical profile.
Fig. 3Diagnostic utility of MYB RNA ISH across the morphological spectrum of adenoid cystic carcinoma (ACC). (A) Minor salivary gland ACC with a prominent tubular architecture raising the differential diagnosis of polymorphous adenocarcinoma in a diagnostic biopsy (H&E, 10×). (B) MYB RNA signals within the tubules and nests indicating a diagnosis of ACC in the biopsy. (C) Recurrent ACC with clear cell pattern (H&E, 10×). (D) Clumped signals of MYB RNA in >30% of the tumour cells at 20× (arrows indicate). (E) Metastatic ACC with signet ring pattern. Focal areas of conventional biphasic cribriform architecture were present. The primary submandibular gland ACC showed a predominance of conventional cribriform pattern with scattered areas of signet ring appearance. (F) Clumped signals of MYB RNA in >30% of the tumour cells (20×). MYB rearrangement was also seen with FISH in this case. (G) ACC with high grade transformation. (H) Most cells showing >10 MYB RNA signals and clumps.
The histology of the eight cases historically diagnosed as ACC that lacked detectable MYB RNA was reviewed. In one of these cases (Fig. 4A), resection of a laryngeal lesion was reclassified as EMC in light of the nodular expansile growth pattern which showed predominance of small tubules lined by luminal cuboidal epithelial and abluminal myoepithelial cells (Fig. 4A). This patient is currently alive without any evidence of disease at 7 years of follow-up. Another case was reclassified as BCAC as it showed a solid nested growth pattern, lacked typical cribriform architecture and showed focal squamous differentiation (Fig. 4B). This patient is also alive without any evidence of disease at 12 years of follow-up.
Fig. 4(A) Resection of a laryngeal lesion reclassified as epithelial myoepithelial carcinoma in light of the nodular expansile growth pattern (H&E, 4×); inset: predominance of small tubules lined by luminal cuboidal epithelial and abluminal myoepithelial cells (H&E, 20×). (B) Parotid gland tumour reclassified as basal cell adenocarcinoma (H&E, 20×). Both cases lacked MYB RNA ISH expression.
The remaining six tumours showed classic morphological features of ACC (Supplementary Fig. 1, Appendix A) with cribriform architecture, invasion into skeletal muscle, and PNI (Supplementary Fig. 1A,B, Appendix A). The majority of these were archival cases with surgeries prior to 2010. The clinicopathological details of the 78 cases of adenoid cystic carcinoma are presented in Table 2.
Table 2Clinicopathological characteristics of the histologically diagnosed adenoid cystic carcinomas with MYB RNA ISH testing results
Thus, ISH for MYB RNA detection had a sensitivity of 92.3% (72/78) for the diagnosis of ACC.
FISH for MYB rearrangement
The clinicopathological details of the patients with ACC that received MYB FISH testing are provided in Table 3. Amongst the 33 cases tested by FISH, MYB rearrangement (Fig. 5) was seen in 14 (42%) cases. All 14 cases with MYB rearrangement by FISH demonstrated MYB RNA by ISH with a 100% concordance. Similarly, six cases without MYB RNA expression also lacked MYB rearrangement by FISH.
Table 3Clinicopathological characteristics of the histologically diagnosed adenoid cystic carcinomas with MYB FISH testing results
Variable
MYB FISH n/median
Positive (14/33)
Negative (19/33)
Age, years (range)
54.3 (19.83–74.75)
49.25 (19.83–67.25)
Gender
Male
5
6
Female
9
13
Type of surgery
Excision
13
18
Biopsy
1
1
Morphological appearance
Conventional tubular or cribriform structures
14
14
Unconventional morphological features
0
3
Solid
0
1
High grade
0
1
Perineural invasion
Yes
7
11
No
7
8
Lymphovascular invasion
Yes
2
1
No
12
18
Tumour max diameter, mm (mean)
9–40 (25)
9–40 (26)
Tumour thickness, mm (mean)
1.5–25 (12)
1.5–12 (8)
Margin
Clear
2
2
Involved
10
17
Close
2
–
Exposed capsule
–
–
Bone invasion
3
2
Lymph node involvement
–
3
Follow-up, years (mean)
1–16.50 (8.5)
1–13.33 (8.5)
Recurrence
Local
1
–
Regional
1
1
Distant
3
4
FISH, fluorescent in situ hybridisation; max, maximum.
Fig. 5(A) Classic morphological appearance of adenoid cystic carcinoma (ACC) (H&E, 10×). (B) MYB RNA ISH positive. (C) Fluorescent in situ hybridisation using a break apart probe for MYB gene locus 6q23.3 in adenoid cystic carcinoma. The red and green fluorophores for the 3′ and 5′ ends of the gene are separated in >15% of the cells indicating MYB gene rearrangement.
The remaining 13 cases demonstrated MYB RNA expression by ISH but lacked MYB rearrangement by FISH (n=11) or failed FISH testing (n=2) (Table 4). Thus, MYB RNA ISH (92.3%) had significantly higher sensitivity for detecting MYB alterations and diagnosing ACC compared with FISH (42%).
Table 4Concordance and discordance of MYB RNA ISH and FISH testing
MYB FISH
Total
Positive
Negative
MYB RNA ISH
Positive
14/33 (a)
13/33 (b)
27 (a+b)
Negative
0/33 (c)
6/33 (d)
6 (c+d)
Total
14 (a+c)
19 (b+d)
33 (n)
A two-by-two table representing concordance/discordance of both diagnostic tests.
Thus sensitivity of FISH test for the diagnosis of ACC was calculated as: (a+c)/n=42.4%.
FISH, fluorescent in situ hybridisation; ISH, in situ hybridisation.
The DNA and RNA extracted from the FFPE samples of the eight cases lacking MYB RNA, including two cases reclassified as EMC and BCAC on review, were sequenced. An FGFR2 (S52W) (COSMIC ID: COSM36903) single nucleotide variation at a variant alle frequency of 20% was seen in a single patient. The other seven patients did not demonstrate any changes identifiable in the 161 gene targeted panel assay.
The RNA extracted from the FFPE did not demonstrate any MYB or MYBL1 fusions in any of the eight cases studied.
The results of the FISH and NGS assays demonstrate that ISH has high sensitivity for detecting increased MYB RNA when present and is a good surrogate marker for MYB gene changes.
MYB RNA expression amongst the mimics of ACC
Of the 58 mimics of ACC included in this study, a single case of BCAC showed MYB RNA expression (1/58). This tumour showed a biphasic pattern with nests showing peripheral palisading of nuclei, areas of squamous and sebaceous differentiation.
All other 57 (98.2%) mimics including PA (Fig. 6A), BCA (Fig. 6B), PAC (Fig. 6C), and EMC (Fig. 6D), were negative for MYB RNA.
Hence, the detection of MYB RNA has 98.2% (57/58) specificity for a diagnosis of ACC amongst salivary gland neoplasms.
Discussion
The current study investigating the diagnostic utility of MYB RNA ISH in 138 cases of primary salivary gland neoplasms with biphasic appearance or cribriform architecture demonstrates a sensitivity of (72/78) 92.3% and specificity of (57/58) 98.2% for a diagnosis of ACC. Rooper et al. report similarly high sensitivity (92%) and specificity (97%) of MYB RNA detection for a diagnosis of ACC.
Furthermore, the results of this study show that in ACC, the sensitivity of detecting MYB RNA by ISH (92.3%) is substantially higher than that of the FISH MYB break apart probe (42%). Further, in the current cohort, concordant MYB RNA ISH expression was observed between primaries and their metastases indicating diagnostic utility of MYB RNA ISH expression in this context. Interestingly, NGS analyses of RNA and DNA extracted from cases where MYB RNA was not detected by ISH also did not show MYB fusion or other genetic changes that could result in increased MYB RNA. This would suggest that RNA ISH is a robust method for the detection of increased MYB RNA, a good surrogate marker for MYB gene alteration, and an extremely useful and efficient diagnostic tool for ACC.
MYB is a nuclear transcription factor that includes the MYBL1 and MYBL2 proteins, which are structurally related and regulate cell proliferation, survival, differentiation and angiogenesis.
Conditional c-myb knockout in adult hematopoietic stem cells leads to loss of self-renewal due to impaired proliferation and accelerated differentiation.
Target genes of V-Myb and C-Myb. Myb transcription factors: their role in growth, differentiation and disease.
in: Frampton J. Myb Transcription Factors: Their Role in Growth, Differentiation and Disease. Proteins and Cell Regulation. Vol 2. Springer Verlag,
Dordrecht2004: 257-270
The activity of MYB is closely regulated in normal tissues. MYB is constitutively activated, and its transcriptional selectivity is altered when its C-terminus is truncated, giving it carcinogenic potential.
The loss of negative regulation, coupled with the juxtaposition of super-enhancer regions of NFIB results in the upregulation of MYB in MYB- NFIB fusion.
have also described fusion of the MYB protooncogene like-1 (MYBL1) and NFIB (MYBL1-NFIB) in a significant number of ACCs without MYB-NFIB fusion. A subset of ACCs without MYB-NFIB fusion also overexpress the MYB protein, through other mechanisms such as alternative RNA splicing.
Increased MYB mRNA transcripts have been detected in 60% of fusion-negative ACC, suggesting that upregulation of MYB mRNA may underlie the pathogenesis in ACC regardless of fusion status.
A critical downstream effect of all these genetic mechanisms is increased production of MYB RNA that can be detected using ISH staining methods.
The major strength of RNA ISH assay is that it can be performed using standard IHC auto-stainers available in most pathology laboratories. RNA ISH assays can be evaluated using routine bright field microscopes which allow assessment of the ISH signals in their histological context. Also, MYB RNA ISH interpretation requires minimal additional assay specific training, making it easy to incorporate into routine diagnostic workflows. Other methods of detecting MYB overexpression, such as IHC are also available. However, MYB IHC has a lower sensitivity at 85% and a significantly lower specificity of 54% for a diagnosis of ACC.
While NGS can be used to identify the genetic changes in MYB, DNA and RNA sequencing are currently significantly more expensive, labour intensive, and not widely available.
The findings of our study and that of Rooper et al.
indicate that detection of MYB RNA ISH is a useful addition to the diagnostic armamentarium for salivary gland neoplasms. The clinicopathological, morphological, and immunohistochemical overlap between pleomorphic adenoma, basal cell adenoma, and ACC is well recognised.
in: Barnes L. Eveson J.W. Reichart P. Sidransky D. World Health Organization Classification of Tumours. Pathology and Genetics of Head and Neck Tumours. IARC Press,
Lyon2005: 221-222
Indeed, all three occur in patients with similar demographic profiles and can show cribriform spaces and basement membrane production. A diagnosis of ACC can be easily made when the tumour is widely infiltrative or shows extensive perineural involvement or has hyperchromatic, angulated nuclei, and easily identifiable mitoses. However, these features may not be readily apparent in fine needle aspiration (FNA) specimens and small biopsies from minor salivary glands. Also, a proportion of ACC can show rounded contours and minimal cytological atypia. In the current study, none of the benign mimics of ACC demonstrated increased MYB RNA. Thus, detecting MYB RNA would carry 100% specificity for a diagnosis of ACC in the clinically critical context of distinguishing ACC from its benign mimics. The combination of morphological features, testing for PLAG1 for pleomorphic adenoma, β-catenin for basal cell adenoma, and MYB RNA for ACC, can be useful in diagnostically challenging cases.
The malignant mimics of ACC include polymorphous adenocarcinoma, basal cell adenocarcinoma, and epithelial myoepithelial carcinoma. Indeed, two cases previously diagnosed as ACC were reclassified as BCAC and EMC on review and lacked MYB RNA. While helpful in distinguishing ACC from these malignant mimics with a specificity of 98.2%, we observed that occasional basal cell adenocarcinoma can demonstrate MYB RNA. Similar findings have also been reported by Rooper et al.
A combination of morphological features, immunohistochemistry for S100, p63, p40, HRAS for epithelial myoepithelial carcinoma, and MYB RNA for ACC can be utilised to distinguish these malignancies from each other.
Polymorphous low grade adenocarcinoma has a consistent p63+/p40− immunophenotype that helps distinguish it from adenoid cystic carcinoma and cellular pleomorphic adenoma.
In the current study, MYB RNA ISH was particularly useful in confirming the diagnosis of ACC in small amounts of tissues sampled from metastatic sites.
Interestingly, six cases morphologically and immunohistochemically diagnosed as ACC both at the time of initial diagnosis and on review did not demonstrate MYB RNA. There were no suggestions of alterations in the MYB gene in these cases by NGS. As shown in Supplementary Fig. 1 (Appendix A) all cases demonstrated a biphasic morphological appearance, cribriform spaces, and basement membrane production and were widely infiltrative into the adjacent soft tissues and skeletal muscle and showed perineural invasion. Interestingly, all of these patients, except one, are alive without disease. The single deceased patient died of unrelated causes. One of these patients demonstrated FGFR2 (S252W) single nucleotide variation. FGFR and NOTCH mutations have been previously described in ACC.
These findings would suggest that the clinicopathological characteristics, particularly survival outcomes of the subset of morphologically diagnosed ACC based on their infiltrative pattern lacking increased MYB RNA, require further analysis. However, most of these tissues lacking MYB alteration by ISH and NGS were from surgeries prior to 2010. While the DNA and RNA extracted from these cases for NGS passed stringent quality checks, degradation of RNA in older cases cannot be entirely excluded.
ISH detection of MYB RNA can be diagnostically useful in ACC. However, it is important to note that a variety of other malignancies, including adenocarcinomas of the colon, breast, and certain leukaemias, can show MYB overexpression. Thus, all evaluations of MYB RNA ISH should include correlation with clinical history, morphological and immunohistochemical features. Indeed, in the head and neck, multiple malignancies including human papillomavirus (HPV)-related multiphenotypic sinonasal carcinomas can demonstrate MYB RNA expression.
In this context HPV33-specific RNA ISH or polymerase chain reaction can be useful as HPV integration is seen in HPV related multiphenotypic sinonasal carcinomas but not in ACC. Similarly, sinonasal undifferentiated carcinomas (SNUC) can show MYB RNA expression,
limiting its utility in distinguishing ACC with high grade transformation from SNUC.
In this study, we observed that suboptimal tissue fixation and decalcification can impact ISH signal identification. The problems were two-fold, either the signals were weak and pale or the slides showed granular dusty background. The increasing use of molecular testing makes it critical that the tissue fixation protocols are optimised to allow molecular diagnostic and predictive techniques. Decalcification of bony tissues should be performed with formic acid-based solutions or chelators such as EDTA.
MYB RNA ISH testing may be utilised on fine needle aspiration cell blocks similar to other routinely used immunohistochemical stains. Ideally, cell blocks should be prepared using the plasma thrombin clot method and fixed in neutral buffered formalin. Reagents containing acetic or picric acid should be avoided. While beyond the scope of this manuscript, a future study evaluating the sensitivity, specificity and the pitfalls in deploying MYB RNA testing in FNA cell blocks will be useful.
Conclusion
In summary, our study provides validation of the findings described by Rooper et al.
using an international cohort of 138 cases of tumours with a prominent cribriform morphology. Detection of MYB RNA has 92.3% sensitivity and 98.2% specificity for a diagnosis of ACC amongst salivary gland neoplasms. ISH for MYB RNA has the highest sensitivity and specificity for ACC amongst various diagnostic tests including IHC and FISH. Furthermore, ISH for MYB RNA can be performed using the standard IHC platforms and evaluated using brightfield microscopy, making it a robust time and potentially cost-efficient diagnostic tool.
Ethics approval
Royal Prince Alfred Hospital Human Research Ethics Committee (2019/ETH07613).
Conflicts of interest and sources of funding
The authors state that there are no conflicts of interest to disclose.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
El-Naggar A.K. Chan J.K.C. Grandis J.R. Takata T. Sootweg P. World Health Organization Classification of Head and Neck Tumours. 4th ed. IARC Press,
Lyon2017e202
Conditional c-myb knockout in adult hematopoietic stem cells leads to loss of self-renewal due to impaired proliferation and accelerated differentiation.
Target genes of V-Myb and C-Myb. Myb transcription factors: their role in growth, differentiation and disease.
in: Frampton J. Myb Transcription Factors: Their Role in Growth, Differentiation and Disease. Proteins and Cell Regulation. Vol 2. Springer Verlag,
Dordrecht2004: 257-270
in: Barnes L. Eveson J.W. Reichart P. Sidransky D. World Health Organization Classification of Tumours. Pathology and Genetics of Head and Neck Tumours. IARC Press,
Lyon2005: 221-222
Polymorphous low grade adenocarcinoma has a consistent p63+/p40− immunophenotype that helps distinguish it from adenoid cystic carcinoma and cellular pleomorphic adenoma.
00084-3&id=gr1.jpg){kind=link}
00084-3&id=gr2.jpg){kind=link}
00084-3&id=gr3.jpg){kind=link}
00084-3&id=gr4.jpg){kind=link}
00084-3&id=gr5.jpg){kind=link}
00084-3&id=gr6.jpg){kind=link}