NUT Carcinoma in a Patient with Unusually Long Survival and False Negative FISH Results

Abstract

Nuclear protein in testis (NUT) carcinoma is a rare and highly aggressive epithelial malignancy defined by rearrangement of the NUTM1 gene on chromosome 15q14. Histologically, NUT carcinoma is an undifferentiated carcinoma formed by sheets and nests of primitive and monotonous “round blue cells” with foci of abrupt keratinization in a subset. NUT carcinoma runs a fulminant clinical course and is almost always quickly lethal, with a median overall survival of only 6.7 months. There is no consensus regarding treatment for this disease, and most patients respond poorly to conventional chemotherapy and radiation. We report a case of NUT carcinoma in an African-American man who initially presented in 2009 with a tracheal mass at age 28. Although fluorescence in situ hybridization (FISH) assays for NUTM1 and BRD4 rearrangements were negative, he was diagnosed based on diffusely positive NUT immunostaining and BRD4-NUTM1 on RNA sequencing. Since his initial presentation, he has undergone multiple surgical procedures and radiation therapy. His tumor has recurred twice, but he has survived for 129 months and is currently alive without disease. Long-term survival of patients with NUT carcinoma is incredibly unusual, especially in patients with tumors that exhibit a BRD4 rearrangement. False negative FISH is a pitfall in diagnosing NUT carcinoma; NUT immunostaining and RNA sequencing are more sensitive diagnostic methods.

Introduction

An uncommon and aggressive tumor, nuclear protein in testis (NUT) carcinoma originates primarily in the thoracic and head and neck regions and is defined by translocation of the NUTM1 gene on chromosome 15 [1]. Typically affecting young adult and pediatric populations, and with a predilection for arising in midline structures, it is a malignant small round cell neoplasm with frequent focal squamous differentiation. Because of its poor response to radiation and standard chemotherapy regimens, and its lack of expression of checkpoint immunotherapy markers [2], median survival is typically only several months for adults, and only slightly longer in children [3–6].

NUT carcinoma bears histologic resemblance to many other tumor types and, even in light of its recent recognition as a distinct entity, still might easily be misdiagnosed. Histologically, NUT carcinoma is comprised of sheets of primitive, round, often discohesive cells with abrupt squamous differentiation in about one-third of cases, mimicking Ewing sarcoma (especially the adamantinoma-like variant) and poorly differentiated squamous cell carcinoma. There is also histologic overlap with sinonasal undifferentiated carcinoma (SNUC), EBV-associated nasopharyngeal carcinoma, olfactory neuroblastoma, pancreatoblastoma, and thymic carcinoma, among many others, depending on the affected site [1, 7–9].

The defining feature of NUT carcinoma is a chromosomal translocation, with the NUTM1 gene fused with one of several partners, resulting in overexpression of the nuclear protein in testis (NUT), easily detectable by immunohistochemistry [10, 11]. Diffuse (> 50%) immunohistochemical positivity for NUT in a speckled nuclear pattern is highly sensitive (87%) and 100% specific, and is therefore considered diagnostic for this tumor [11]. The most frequent translocation, BRD4-NUTM1, accounts for 70% of cases. Other gene fusions include BRD3-NUTM1 and NSD3-NUTM1 [5, 12]. NUT fusion products are theorized to lead to neoplasia in two ways: they drive tumor growth, and they maintain cells in an undifferentiated state [8, 13]. Cellular differentiation is blocked when the bromodomains of NUT fusion products bind to acetylated histones in association with p300 proteins [14].

As patients with NUT carcinoma show poor responses to standard radiation and chemotherapy, these molecular mechanisms have proven particularly translationally relevant, as their elucidation has opened new avenues for treatment. Bromodomain and extra-terminal (BET) inhibitor drugs are acetyl-histone mimetics that competitively inhibit the binding of fusion products such as BRD4-NUTM1, and clinical trials have demonstrated some efficacy in treatment of NUT carcinoma [15–17]. Additionally, targeting the histone deacetylase (HiDAc) enzyme with an inhibitor results in promoting overall histone acetylation and promoting differentiation to a squamous phenotype, and has seen clinical effect as well [16–19]. Despite these therapeutic successes, patient survivorship of NUT carcinoma is still dismal, with a 5-year survival rate of around 7% [20]. Exceptional cases have been reported with survival as long as 35, 47, 72, and 78 months after diagnosis [4, 18], with rare reports of survival 108 and 144 months after diagnosis [21, 22]. Here, we present the case of a 28-year-old male patient who was diagnosed with tracheal NUT carcinoma with a BRD4-NUTM1 fusion on RNA sequencing after negative FISH testing. The patient has survived his disease for 129 months, and is still alive with no evidence of disease at the end of the follow-up period.

Case Report

A 28-year-old African-American male presented to UT Southwestern Medical Center in September, 2009 with a chief concern of 2-year history of progressive difficulty breathing and a tracheal mass. He did not use tobacco products or abuse alcohol. A computed tomography (CT) scan demonstrated an 8-mm soft tissue mass affecting the trachea and subglottic area at the level of the thyroid gland (Fig. 1). The patient underwent microsuspension laryngoscopy and excision of the tracheal mass with laser ablation. Intraoperatively, it was noted that the mass appeared papillary and was located on the left posterolateral trachea near the cricotracheal junction.

Fig. 1.

Computed tomography (CT) scan from 2009 reveals a small intraluminal tracheal soft tissue mass (arrows)

The excisional specimen was submitted for histologic evaluation, showing a small, superficial epithelial neoplasm intimately growing with submucosal seromucinous glands (Fig. 2a). Focally, vague squamoid features were noted (Fig. 2b). Mild nuclear atypia was present, but mitotic figures were infrequent. Because of the tumor’s association with seromucinous glands, a diagnosis of “low-grade adenocarcinoma” was initially made, with a listed differential diagnosis that included salivary-type adenocarcinomas, amongst other entities. The patient returned to the operating room in December 2009 to undergo cricotracheal resection for re-excision of a margin with concern for disease recurrence. The excisional specimen demonstrated histologic features similar to those of the previous surgical material. At that time, the case was sent to two outside pathologists, and the diagnosis was revised to “invasive carcinoma with features of non-keratinizing/basaloid squamous cell carcinoma, involving seromucinous glands.” The margins of the resection specimen were free of tumor. The patient was followed by head and neck surgery for regular tumor surveillance until August 2014 (56 months after initial resection), when a tracheal mass at the left infraglottic region was noted on bronchoscopy. The patient again returned to the operating room for excision of the mass, and the specimen showed recurrent carcinoma with histologic features identical to the previous surgical material.

Fig. 2.

Histopathologic characteristics of the initial tumor resection, in 2009. a Hematoxylin and eosin (H&E) stain reveals invasive islands of a malignant small round blue cell tumor (original magnification ×10). b At higher power, tumor nuclei are primitive and monotonous; focal pseudoglandular structures and mucous cells are noted (H&E, original magnification ×20)

Tumor surveillance continued for 59 additional months until July 2019, when the patient presented with a concern of vocal hoarseness. Left true vocal cord paralysis was noted, and a CT scan of the neck demonstrated a large nodule present in the left lobe of the thyroid gland. Thyroid ultrasound and fine needle aspiration (FNA) were performed, and the FNA revealed malignant cells in a polymorphous lymphoid background. Immunohistochemistry (IHC) for p40 was positive, while stains for TTF-1, PAX-8 and PTH were negative. The patient underwent a left thyroid lobectomy and level 6 neck dissection. The resection specimen showed a grey-white lobular mass of tumor that measured 5.2 cm (Fig. 3). It was noted that the tumor was centered within the perithyroidal soft tissue; involved the left recurrent laryngeal nerve; and invaded the thyroid gland. Histologically, the tumor consisted of sheets and islands of primitive, basaloid cells (Fig. 4a). Foci of abrupt keratinization were seen at the centers of many tumor islands (Fig. 4b). Mitotic activity was significantly elevated (Fig. 4c), and tumor necrosis was identified. Lymphovascular invasion was seen, and metastatic carcinoma was present in 2 of 7 paratracheal lymph nodes. IHC evaluation for NUT was diffusely positive in a speckled nuclear pattern (clone c52b1, Cell Signaling) (Fig. 4d), and p40 was positive as well, while CD99, CD5, TTF-1, synaptophysin, chromogranin and CD34 were negative. The tumor was deemed to be histologically identical to the previous specimens.

Fig. 3.

Gross pathology from the 2019 resection revealed a large, solid, and lobulated tan-white mass abutting (left) and invading (right) the beefy red thyroid gland parenchyma

Fig. 4.

Histopathologic features of the 2019 resection specimen. a Discohesive sheets of primitive small round blue cells comprise much of the malignancy (H&E, original magnification ×10). b Abrupt keratinization and tumor necrosis are evident (H&E, original magnification ×20). c Tumor cell nuclei are large and round, with prominent nucleoli; increased mitotic figures are present (H&E, original magnification ×40). d IHC for NUT protein demonstrates a speckled nuclear pattern of positivity in tumor cells (original magnification ×20)

Dual color bring-together (to test for a BRD4-NUTM1 fusion) and split-apart (to test for rearrangement of NUTM1) FISH on 5 micron FFPE sections of tumor was performed as described [23]. FISH probes used were as follows: NUTM1: 5′ centromeric probes, RP11-368L15 and RP11-1084A12 (biotin labeled, red) and 3′ telomeric probes, RP11-1H8 and RP11-64o3 (digoxigenin labeled, green); BRD4: 5′ centromeric probes, RP11-207i16 and RP11-3055m5 (biotin labeled, red), and 3′ telomeric probes, RP11-319O10 and RP11-681D10 (digoxigenin labeled, green). To determine whether a BRD4-NUTM1 fusion was present, dual color FISH using the 3′ telomeric NUTM1 and 5′ centromeric BRD4 probes were used; whereas both 5′ centromeric and 3′ telomeric probes were both used for NUTM1 to detect rearrangement of the respective loci. 200 nuclei were counted in four different areas of each tumor. All FISH assays were negative for fusions of NUTM1 and BRD4.

Targeted RNA sequencing (RNA-seq) as described previously [24]. Briefly, whole-slide tissue sections were cut at 10 µm, and Qiagen AllPrep kits (Qiagen, Germantown, MD) were used for RNA isolation. A sequencing library was generated using a modified TruSight RNA Pan-Cancer kit (Illumina, San Diego, CA) with 1425 genes. Sequencing was performed on the NextSeq 550 (Illumina, San Diego, CA) with a minimum of 6 million mapped reads. Fusions were called using the Star-Fusion algorithm [25], and were manually reviewed via the Integrated Genomics Viewer (Broad Institute, Cambridge, MA). RNA-seq was positive for BRD4-NUTM1 gene fusion (BRD4ex10-NUTM1ex3).

The final diagnosis was NUT carcinoma. The patient has since been treated with radiation therapy in addition to his surgical procedures. He is alive free of disease as of June 2020, having survived for a total of 129 months.

Discussion

NUT carcinoma is a particularly aggressive small round cell tumor which carries a very poor prognosis, and it is frequently mistaken for other tumor types. General awareness of this entity is gaining momentum, in large part due to the availability of an immunohistochemical assay highly sensitive and specific for aberrant expression of the NUT protein [11]. Although novel therapies based on epigenetic tumor survival mechanisms are being developed, overall prognosis remains dismal as these tumors respond poorly to standard chemoradiation regimens [20].

Having undergone several surgical resection procedures, as well as radiation therapy, the patient we report has survived a total of 129 months since resection of the original, 8 mm tracheal mass in 2009. Histologically (as does a typical case of NUT carcinoma), this tumor showed consistently unfavorable features, including necrosis, frequent mitoses, lymphovascular invasion, as well as metastatic deposits in regional lymph nodes. Additionally, an even more abysmal prognosis was portended by the presence of a NUTM1 fusion with BRD4 in this tumor, the least favorable known rearrangement [26]. Our patient’s curiously long survival would suggest something biologically unique is occurring, intrinsic either to his tumor or to the patient himself.

Most patients diagnosed with NUT carcinoma have a life expectancy of only several months [4, 8, 22]. The International NUT Midline Carcinoma Registry (INMCR) defines long-term survival of this disease process as 3 years’ survival from time of diagnosis. Chau et al. [26] recently reviewed 141 cases of NC that were reported in the INMCR in the largest review of NC cases to date. Out of those patients, just 16 survived at least 3 years, six patients survived at least 5 years, and only one patient survived at least 10 years [26]. Another review by Lemelle, et al. [22] documented a case of a 49-year-old male patient who survived over 12 years and was alive at the end of the follow-up period.

Chau et al. [26] identified within their cohort three statistically-distinct risk groups of NC patients based upon anatomic site involved and tumor genetics. Site was subdivided into thoracic (51% of cases) and nonthoracic (49%) locations. BRD4-NUTM1 fusions were the most common genetic alteration, identified in 78% of those patients, followed by BRD3-NUTM1 fusions (15% of patients) and NSD3-NUTM1 fusions (6%). The risk stratification groups include:

1. Group A: nonthoracic primary; BRD3- or NSD3-NUT fusion.

2. Group B: nonthoracic primary; BRD4-NUT fusion.

3. Group C: thoracic primary; any NUT fusion.

According to the system proposed by Chau et al. [26], our patient falls into Group B, with a nonthoracic primary and BRD4-NUT fusion. Long-term survivors have been identified in this group, but patients with non-BRD4-NUT fusions showed better survivorship in their review. Group C, into which patients with thoracic primaries are placed, show the worst overall survival, regardless of the gene fusion present.

Explanations for extraordinarily long survival in NC patients are elusive at best. It is possible that genetic or epigenetic dynamics, beyond the mere presence of a fusion gene or overexpression of the NUT protein, are at work. However, the typical aggressive histopathologic features of this tumor and classic fusion (BRD4ex10-NUTM1ex3) suggests the reason for our patient’s resilience to lie not in features of the tumor, but rather in features of the patient himself. This patient was not alone among those with NC in his relative youth at disease onset, and freedom from comorbidities that might contribute to poorer prognosis in older populations with similarly aggressive lesions. Most NC cases arise in and around midline structures, so a tracheal location is not uncommon; but perhaps specific anatomic features may have been contributory in this case (e.g., early symptoms due to airway involvement, intimate growth within submucosal glands). It is possible that some unique and uncharacterized cellular milieu may have contributed to a tumor microenvironment in which growth was somehow slowed or metastatic potential mitigated, and it is unknown to what extent race or ethnicity might play a role. Overall, a credible explanation is not yet forthcoming but, as more cases are reported and more data accumulated, studies may be undertaken to further refine the contingent of prognostic indicators.

Finally, this case highlights the point that FISH is not entirely specific for a diagnosis of NC, and a negative result should not be regarded as definitive for ruling it out [11]. It also underscores the superior sensitivity and specificity of NUT IHC. Indeed, in our collective experience, we have never encountered a case with > 50% staining with NUT IHC that was not ultimately proven to be NC (unpublished observations). The reason for false-negative FISH results in some NC cases is not entirely clear, but it may be a result of the fact that many of the NC translocations arise from chromoplexy, where there are up to thirty rearrangements from a single catastrophic event that result in the single oncogenic fusion [27]. The takeaway points are that NUT immunoexpression is diagnostic for NC, and if genetic confirmation or documentation of the precise fusion is desired, a sequencing (e.g., next generation sequencing) modality is preferred over FISH.

Funding

This research was funded by the Jane B. and Edwin P. Jenevein, MD Endowment for Pathology at University of Texas Southwestern Medical Center.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflicts of interest to disclose.