A and B, Micrographs of an intermediate-grade MEC of the parotid with extensive fibrosis and unusual keratinization, raising the possibility of metastatic squamous cell carcinoma on fine-needle aspiration (hematoxylin-eosin, original magnification ×10 and ×20, respectively). C, FISH confirmed the histologic impression of MEC cells with the MAML2 gene rearrangement showing 1 pair of green and orange signals in juxtaposition and 1 green signal and 1 orange signal separately. D and E, Sclerosing MEC, a rare variant of thyroid MEC, with eosinophilia. The histologic evaluation shows a fibrotic background, abundant eosinophilic infiltrate, and conventional elements of MEC, such as mucinous and intermediate cells (hematoxylin-eosin, original magnification ×10 and ×20, respectively). F, FISH confirmed the presence of MAML2 gene rearrangement in keeping with the morphologic impression of MEC.
Survival of patients with mucoepidermoid carcinoma was stratified by CRTC1/MAML2 translocation status (present or absent). Overall and disease-free survival are given by translocation status. Vertical lines indicate 5 years.
Rami E. Saade, Diana Bell, Joaquin Garcia, Dianna Roberts, Randal Weber. Role of CRTC1/MAML2 Translocation in the Prognosis and Clinical Outcomes of Mucoepidermoid Carcinoma. JAMA Otolaryngol Head Neck Surg. 2016;142(3):234–240. doi:10.1001/jamaoto.2015.3270
The CRTC1/MAML2 fusion transcript, which arises from the CRTC1/MAML2 translocation, is a molecular marker unique to mucoepidermoid carcinoma (MEC), the most common malignant tumor of the salivary gland. The extent to which the transcript influences disease features and patient survival is unclear.
To determine whether the CRTC1/MAML2 fusion transcript is associated with disease stage, tumor grade, or survival outcomes in patients with MEC.
Design, Setting, and Participants
A retrospective medical record review was performed at a tertiary-care academic medical institution. The review included 90 patients with MEC who underwent treatment from January 1, 1995, to December 31, 2011, and for whom archived formalin-fixed, paraffin-embedded tumor specimens were available. Records were reviewed for clinical, demographic, and survival data. Tumor specimens underwent fluorescence in situ hybridization. Follow-up was completed on May 15, 2014, and data were analyzed from June 1 to July 1, 2014.
Main Outcomes and Measures
CRTC1/MAML2 fusion transcript status. Statistical analysis determined whether transcript status was associated with disease stage, tumor grade, and/or overall and disease-free survival.
Among the 90 eligible patients (median [range] age, 55.1 [7.8-89.2] years), 42 were female and 48 were male. Fluorescence in situ hybridization revealed a CRTC1/MAML2 translocation in 50 patients (56%). The translocations were more prevalent in intermediate-grade tumors (31 of 49 [63%]) than in high-grade (11 of 49 [22%]) and low-grade (7 of 49 [14%]) tumors; 1 tumor sample had no available grading. Similar proportions of patients with translocation-positive disease had T1 (13 of 49 [26%]), T2 (15 of 49 [31%]), T4a (14 of 49 [28%]), or T0 or Tx (8 of 49 [16%]) stages of disease. Thirty-eight of 49 patients with translocation-positive MEC (78%) had N0 stage of disease. Rates of 5-year overall survival were similar for patients with translocation-positive and translocation-negative disease (76.8% vs 75.5%, respectively; P = .17), as were rates of disease-free survival (65.2% vs 57.4%, respectively; P = .28).
Conclusions and Relevance
Detection of the CRTC1/MAML2 fusion transcript provides useful information for MEC diagnosis but is not associated with differences in survival outcomes.
Mucoepidermoid carcinoma (MEC), the most common malignant tumor of the salivary tissue, arises from major and minor salivary glands along the aerodigestive tract and tracheobronchial tree. Clinical and pathologic markers have been used to determine the disease’s behavior and biological features, and patient age, tumor size, the presence of cervical lymphadenopathy, distant spread, perineural involvement, and histologic grade have been used to stratify patients and determine treatment. Low-grade, low-stage tumors typically require only surgical resection, whereas high-grade, high-stage tumors necessitate multimodality therapy that includes adjuvant radiotherapy and/or elective neck dissection. Conventional classification of MECs include low-, intermediate-, or high-grade tumors depending on the extent to which they exhibit adverse features, such as perineural invasion, angiolymphatic invasion, coagulative necrosis, and infiltrative growth; a high mitotic rate and a cystic component of less than 20% are indicative of higher-grade tumors. However, because of the substantial cellular heterogeneity of these tumors and the significant variability in interpreting and scoring their grades, reproducibly grading MECs in a way that has prognostic value remains challenging.1,2
Specific molecular signatures and chromosomal aberrations could play a pivotal role in identifying high-risk groups of patients with MEC. Chromosomal aberrations are characteristic of neoplasia. Recurrent, nonrandom chromosomal translocations and the creation of novel chimeric fusion oncogenes are well recognized and causally implicated in carcinogenesis. Tumor-specific chromosomal rearrangements often produce potent fusion oncogenes, which induce tumorigenesis by deregulating the cell cycle, which results in the overexpression of a gene in 1 of the breakpoints, or by causing the fusion of 2 genes, 1 in each breakpoint, which results in a hybrid, chimeric gene.3 At present, almost 400 critical gene fusions have been identified in human cancers; these oncogenes account for 20% of human cancers.4 Fusion oncogenes are often derived from and encode for transcription factors, transcription regulators, and receptor tyrosine kinases, which frequently are involved in oncogenesis. Among such aberrations, the t(11;19)(q12;p13) translocation is unique to MEC and was first described by Nordkvist et al5 in 1994 and characterized by Tonon et al6 in 2003. The recurrent translocation that involves the CRTC1/MAML2 genes (GenBank AY040324) has been identified as the underlying genetic event in most MECs.
Several studies7- 9 have used reverse transcription polymerase chain reaction and fluorescence in situ hybridization (FISH) to detect the CRTC1/MAML2 fusion transcript in more than 55% of patients with MEC. In those studies, patients with translocation-positive disease had significantly better survival than those with translocation-negative disease, which suggests that the CRTC1/MAML2 transcript has prognostic significance in MEC. Behboudi et al8 reported that patients with translocation-positive and translocation-negative MEC had median survival times of 10.0 and 1.6 years, respectively, and that those with translocation-positive MEC also had a significantly lower risk for local recurrence, metastasis, and tumor-related death than patients with translocation-negative MEC. However, Seethala et al2 later found that although patients with translocation-positive MEC had better disease-specific survival, the disease-free survival of patients with translocation-positive and translocation-negative MEC did not differ significantly. In view of the high rate of detection of the CRTC1/MAML2 fusion transcript in MECs and its subsequent effect on diagnosis, prognosis, and treatment, we evaluated the data of a cohort of 90 patients with MEC with available fusion transcript status.
We identified 90 patients with MEC treated at The University of Texas MD Anderson Cancer Center from January 1, 1995, to December 31, 2011, for whom archived formalin-fixed, paraffin-embedded tumor specimens were available from the department’s tissue bank. Pathologic patterns and CRTC1/MAML2 fusion transcript status were recorded independently, and we assessed differences in clinical factors, including sex, age, and disease stage, and differences in clinical outcomes between patient groups based on pathologic pattern or fusion transcript status for significance. The study was approved by the institutional review board of the MD Anderson Cancer Center. For this retrospective medical record review, the data were deidentified and the institutional review board determined that informed consent was not required.
We used FISH to determine each patient’s CRTC1/MAML2 fusion transcript status. Two 1.0-mm-diameter tissue cores from different areas of each patient’s formalin-fixed, paraffin-embedded MEC tissue sample were used to create tissue microarrays. The microarrays were subjected to dual-color FISH, which was performed using a MAML211q21 bacterial, artificial chromosome, break-apart probe and the clones RP11-16K5 and RP11-676L3 (labeled spectrum orange and spectrum green, respectively), as described previously.10 At least 60 cells in the target regions then underwent analysis using fluorescence microscopy. Normal cells without the MAML2 gene rearrangement show green and orange signals in juxtaposition, which typically result in a single yellow conglomerate signal; in contrast, cells with the MAML2 gene rearrangement show 1 pair of green and orange signals in juxtaposition and 1 green signal and 1 orange signal separately. Specimens in which the split signal was identified in more than 20% of the nuclei based on internal validations of normal salivary gland were considered positive for the MAML2 rearrangement (ie, having the CRTC1/ MAML2 fusion transcript). Negative control specimens consisted of normal salivary gland tissue; positive control specimens consisted of H3113 and H292 cell block sections and MEC specimens that had been identified previously as carrying the MAML2 translocation. FISH studies show only MAML2 gene alterations; they do not indicate the fusion partner or entirely exclude inversions. FISH was also performed on full sections for 20 tumors (these were included on the tissue microarrays) (exemplified in Figure 1).
The MECs were graded as low, intermediate, and high. The grades were reviewed separately by 2 dedicated head and neck pathologists (D.B. and Adel el Naggar, MD, PhD).
Data were analyzed from June 1 to July 1, 2014. Descriptive statistics for scaled values and frequencies of study patients within the categories for each of the biomarkers of interest were enumerated with the assistance of the commercial statistical software listed below. Correlations between biomarkers and between biomarkers and end points were assessed using the Pearson χ2 test or, for fewer than 10 participants in any cell of a 2 × 2 grid, the 2-tailed Fisher exact test. Overall survival was defined as the time from presentation to death or last contact; disease-free survival was defined as the time from the end of treatment for initial disease to death or last contact. Curves describing overall and disease-free survival were generated using the Kaplan-Meier product-limit method, and differences between the actuarial curves were assessed for significance using the log-rank test. Follow-up was defined as the time from the first appointment for evaluation of the primary tumor until the last contact or death and was completed on May 15, 2014. P < .05 was considered significant. All statistical tests were performed using the Statistica (StatSoft, Inc) and SPSS Windows (SPSS, Inc) statistical software applications.
Patient characteristics are given in Table 1. Eighty-seven patients (97%) underwent preoperative diagnostic fine-needle aspiration cytology, which indicated malignant tumor in 65 patients (75%) and benign processes in 19 (22%) but yielded a noncontributory pathologic process for 3 patients (3%). In terms of treatment, 40 patients (44%) had surgery alone; 2 (2%), radiotherapy alone; 40 (44%), surgery and radiotherapy; and 8 (9%), surgery, radiotherapy, and chemotherapy. Forty-eight of 50 patients (96%) who received radiotherapy did so in the postoperative adjuvant treatment setting. Among the 8 patients who received chemotherapy, only 1 (13%) had definitive chemotherapy; most patients received neoadjuvant (4 [50%]), adjuvant (2 [25%]), or neoadjuvant and adjuvant (1 [13%]) chemotherapy. Among the 88 patients who underwent surgery, 41 (47%) underwent a neck dissection. The margins of the surgical specimens for patients undergoing surgery at an outside institution and those undergoing surgery at MD Anderson Cancer Center were clear in 62 of 88 cases (70%). Of the 90 patients included in the study, 68 (76%) were alive at least 5 years after completing treatment. Disease characteristics, including TNM stage distribution, grade distribution, anatomic site distribution, and other tumor features such as perineural invasion status are given in Table 1.
FISH was performed on the 90 tumor specimens. Figure 1 illustrates examples of FISH findings for full sections of paraffin-embedded tissue with histologic findings for sclerosing MEC.
FISH detected a CRTC1/MAML2 translocation in 50 of the 90 tumor specimens (56%); of these, 49 patients had an available histologic grade. Among the 49 patients with available histologic grading and a CRTC1/MAML2 translocation, 31 (63%) had intermediate-grade tumors, only 11 (22%) had high-grade tumors, and 7 (14%) had low-grade tumors (Table 2). With respect to T stage, the 50 CRTC1/MAML2 translocations were similarly distributed across patient groups: 13 patients (26%) had T1 disease; 15 (31%), T2 disease; 14 (28%), T4a disease; and 8 (16%), T0 or Tx disease. Thirty-eight patients (78%) with translocation-positive findings had N0 disease.
The 5-year overall survival rates of the patients with translocation-positive MEC (76.8%) and translocation-negative MEC (75.5%) did not differ significantly (P = .17; Figure 2). The 5-year disease-free survival rates of the translocation-positive patients (65.2%) and translocation-negative patients (57.4%) also did not differ significantly (P = .28; Figure 2).
Previous reports of CRTC1/MAML2 translocation positivity by tumor grade in MEC2,8,9,11- 13 are summarized in Table 3. Despite its flaws, fine-needle aspiration cytology is used increasingly as a part of the diagnostic triage for salivary gland tumors. With respect to MEC, previous studies14 suggest that fine-needle aspiration cytology is diagnostically accurate in high- or intermediate-grade tumors but unsatisfactory for low-grade tumors. Our study’s findings lend further support to the notion that the application of molecular techniques to cytologic material to detect the CRTC1/MAML2 fusion transcript and/or protein may be helpful in cases of uncertainty (although clinical studies are required to validate such an approach). Although early studies8,12 reported that the fusion transcript was restricted to low- and intermediate-grade MEC, the present study and others9,11 have detected the translocation in high-grade MEC, albeit at rates lower than in low- and intermediate-grade MEC. Nevertheless, the finding that high-grade MEC also expresses the fusion transcript suggests that the detection of the transcript may be helpful in distinguishing MEC from poorly differentiated adenocarcinoma or clear cell carcinomas when conventional histologic distinction is difficult. Toward the other end of the spectrum, low-grade variants of MEC that do not express the transcript could be misclassified as benign disease, such as Warthin tumor, oncocytoma, or cystadenoma.2 Oncocytic and mucinous metaplasia are notorious for being present in various salivary neoplasms.
Although patient age, disease stage, and tumor grade are proven MEC prognosticators, even histologically low-grade MECs have been anecdotally reported to behave aggressively,2 which underscores the difficulty of accurately determining the prognosis and management of MEC. The most challenging category of MEC in terms of determining prognosis and management is intermediate-grade MEC. No universal agreement exists regarding which histologic grading criteria are the most useful, and grading has varied.2 Thus, additional tests to help determine the aggressiveness and consequent treatment of MEC are needed. Several previous studies8,9,12 have shown that the CRTC1/MAML2 translocation can be regarded as a biomarker of favorable prognosis, and the detection of the translocation in a patient with MEC may influence the decision to perform neck dissection or to recommend radiotherapy. Although retrospective data support the value of the translocation status in determining MEC prognosis, other retrospective studies suggest that the translocation status has no bearing on prognosis. For instance, Seethala et al2 showed that the translocation was not a powerful predictor of outcome in a cohort of 48 patients with MEC. This finding is in agreement with those of our larger present study, in which the 5-year overall survival and disease-free survival rates of patients with translocation-positive and -negative MEC did not differ significantly.
Retrospective reviews carry inherent limitations, including different patient presentations, different therapies, and different available outcome measures. These data must be validated prospectively. When we look more closely at the prior studies (Table 3) and compare patient and disease presentations, stage and grade distributions, methods, and designs, no major culprit explains the opposing results on the prognostic value of CRTC1/MAML2 except for our larger cohort (90 specimens). Whether patients presenting to our institution with recurrent disease introduced a patient population with highly refractory MEC into the retrospective study is debatable. Although this possibility could represent a valid pitfall in our study, we believe that subtracting 14 patients from the cohort would weaken the statistical power. Therefore, we did not perform a subgroup analysis. Underlying confounders may be covering a positive effect of the CRTC1/MAML2 translocation on outcome and prognosis. However, a multivariate analysis could not be performed owing to technical shortcomings that relate to the number of variables and statistically significant biomarkers.
Other authors7,11,15 have suggested that different aberrations could be used alongside the CRTC1/MAML2 translocation to determine prognosis in patients with MEC. Anzick et al7 identified CDKN2A (GenBank X94154) deletions and methylation in patients with CRTC1/MAML2 translocation-positive MEC and poor prognosis but did not detect CDKN2A deletions or hypermethylation in any patient with good prognosis. These findings seem to indicate that such CDKN2A alterations may underlie tumorigenesis in MEC and/or other malignant salivary neoplasms. The CRTC1/MAML2 and CDKN2A status probably could be used to define more homologous prognostic categories of patients with MEC, a finding that may have biological and therapeutic implications.7 In addition to the t(11;19)(q21;p13) translocation frequently described in MEC, Fehr et al11 described a novel fusion between CRTC3 and MAML2. The CRTC gene family includes human genes CRTC1 (ie, MECT1), CRTC2 at 1q21, and CRTC3 at 15q26; MECT1 has 32% homology with the latter 2 genes.16 In a study of 101 patients with MEC, Nakayama et al15 detected MECT1-MAML2 and CRTC3-MAML2 fusion transcripts in 34% and 6% of patients, respectively. These 2 fusion transcripts were mutually exclusive, and no patients had CRTC2-MAML2 fusion transcripts. The investigators found no significant difference in survival between patients with either translocation, but patients with translocation-positive disease had better disease-free survival than those with translocation-negative disease.15
The presence of the CRTC1/MAML2 fusion transcript in a substantial proportion of patients with MEC and its putative role in tumor initiation and progression suggest that the transcript can be targeted therapeutically. However, in MEC, the CRTC1/MAML2 fusion product is not an enzyme but instead acts in conjunction with other proteins to form transcription activation complexes that act on Notch- and cAMP (cyclic adenosine monophosphate)–response element binding–regulated pathways. Therefore, the best way of therapeutically targeting CRTC1/MAML2 transcript–related processes in MEC is somewhat unclear. To date, Notch inhibition via a number of mechanisms has garnered the most attention. Anti-Notch agents under investigation for clinical use include γ-secretase inhibitors, although the finding that CRTC1/MAML2 transcript–mediated Notch-target activation occurs in the presence of γ-secretase inhibitors suggests that these agents are unlikely to be effective in MEC. More promising are RNA interference strategies aimed at abrogating tumor progression, although these strategies are still investigational.
Mucoepidermoid carcinoma is a malignant tumor of the salivary gland with a heterogeneous biology and spectrum of clinical behavior. Deciphering the molecular signatures of the disease is key to optimizing diagnosis prognostication and potential targeted therapy. Our findings suggest that detection of the CRTC1/MAML2 fusion transcript provides useful information for MEC diagnosis but is not by itself a powerful predictor of outcomes.
Correction: This article was corrected on March 17, 2016, to replace 2 sentences in the Results section.
Corresponding Authors: Rami E. Saade, MD, Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030 (firstname.lastname@example.org).
Submitted for Publication: May 23, 2015; final revision received September 28, 2015; accepted November 11, 2015.
Published Online: January 21, 2016. doi:10.1001/jamaoto.2015.3270.
Author Contributions: Drs Saade and Bell contributed equally to the study. Drs Saade and Bell had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Saade, Bell.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Saade, Bell, Garcia, Roberts.
Critical revision of the manuscript for important intellectual content: Saade, Bell, Garcia, Weber.
Statistical analysis: Roberts.
Obtained funding: Bell.
Administrative, technical, or material support: Saade, Bell.
Study supervision: Bell, Garcia, Weber.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported in part by start-up funds from The University of Texas MD Anderson Cancer Center (MDA) (Dr Bell).
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Adel el Naggar, MD, PhD, Department of Pathology, MDA, helped review and grade the specimens. Rebecca Wehrs, BS, Mayo Laboratories, provided technical assistance with fluorescence in situ hybridization. Josef Munch, BA, Department of Scientific Publications, MDA, helped with editing of the manuscript. None of the contributors received compensation for his or her role.