Clinical and molecular features of pulmonary NUT carcinoma characterizes diverse responses to immunotherapy, with a pathologic complete response case

Abstract

Purpose

Nuclear protein in testis (NUT) carcinoma is an uncommon malignant cancer characterized by NUTM1 rearrangement. We aimed to investigate the clinicopathological and molecular features and immunotherapy of pulmonary NUT carcinoma.

Methods

Immunohistochemistry (IHC) for NUT (C52B1) and programmed cell death ligand 1 (PD-L1: 22C3) and fluorescence in situ hybridization (FISH) for NUTM1 break and BRD4–NUTM1 fusion were performed on six pulmonary NUT carcinoma samples.

Results

The 6 pulmonary NUT carcinoma samples were obtained from 5 males and 1 female, with ages ranging from 31 to 73 years (average, 46 years). Five tumors occurred in the lobes, with one in the trachea. Pathologically, all cases showed primitive-appearing round to epithelioid cells growing in nests and sheets. Squamous differentiation and abrupt keratinization were observed. All tumors expressed the NUT protein and p63, and 4 tumors showed focal synaptophysin, but PD-L1 expression was not observed. All cases displayed NUTM1 rearrangement, 5 had BRD4–NUTM1 fusion, and one had an unknown partner. Three patients presented regional lymph node involvement at diagnosis. Five patients underwent intensive radiation and/or chemotherapy. Furthermore, 2 patients (1 and 2) received a combination of PD-L1 inhibitor and chemotherapy. Patient 1 exhibited a poor response and soon showed tumor progression and metastasis; however, patient 2 responded remarkably and achieved pathologic complete response (pCR) without uncontrollable adverse events. The overall survival time was 2.9 months.

Conclusions

Pulmonary NUT carcinoma exhibits poorly differentiated morphological features with diffuse NUT staining, low PD-L1 expression, and NUTM1 rearrangement. Despite its poor prognosis, it presents a diverse response to immunotherapy. Immune checkpoint inhibitors (ICIs) need to be further explored in NUT carcinoma.

Introduction

Nuclear protein of testis (NUT) carcinoma is a rare, highly aggressive, and newly recognized subtype of poorly differentiated carcinoma (French et al. 2003; Travis et al. 2015). Fewer than 100 cases of NUT carcinoma have been reported (Travis et al. 2015). It usually arises in midline sites, such as the head and neck or thorax, and is occasionally diagnosed in other tissues or organs (Chatzopoulos et al. 2021; French et al. 2014). Genetically, NUT carcinoma harbors NUTM1 gene rearrangement, and most patients have BRD4–NUTM1 fusion (78%), followed by BRD3-NUTM1 (15%) and NSD3-NUTM1 (6%) (Chau et al. 2019; French et al. 2003). It lacks typical histological features, whereas foci of abrupt keratinization are often present (Bauer et al. 2012; Evans et al. 2012; Sholl et al. 2015). Diffuse staining of NUT via immunohistochemistry (IHC) and molecular detection help to differentiate NUT carcinoma from other poorly differentiated carcinomas (Haack et al. 2009).

NUT carcinoma is lethal. A large proportion of NUT carcinomas present with lymph node involvement (75%) and metastases (67%) at diagnosis (Bauer et al. 2012). A multimodal approach with systemic surgery, chemotherapy, and radiotherapy has been adopted in clinical practice for NUT carcinoma. Of note, several target agents have emerged as candidates for treatment. Even with intensive treatment, patients with this disease exhibit a poor response, and the median overall survival (OS) is 6.5 months, with the median OS of primary pulmonary NUT carcinoma being only 4.4 months (n = 67, from 3.5 to 5.6 months) (Chau et al. 2019).

Recently, immune checkpoint inhibitors (ICIs) have resulted in dramatic tumor reductions in patients with solid cancers such as advanced lung cancer (Mansfield et al. 2020; Antonia et al. 2016), hepatocellular carcinoma (Yang et al. 2020), renal cell cancer (Rini et al. 2019), and other rare tumors (Wang et al. 2020, 2021). Programmed cell death 1 (PD-1)/PD-L1 inhibitors can block the immune checkpoint pathway and restore the killing effect of immune cells against tumor cells. Immunotherapy drugs include PD-1 inhibitors (nivolumab and pembrolizumab) and PD-L1 inhibitors (atezolizumab and durvalumab). The vast majority of patients receive immunotherapy as second-line or concurrent treatment. To date, there are few case reports on the use of immunotherapy in patients with pulmonary NUT carcinoma (Xie et al. 2020; Cho et al. 2020; Hung et al. 2021; Maruyama et al. 2018; Davis et al. 2021; Joel et al. 2020; Riess et al.2021; Gupta et al. 2022; Zhou et al. 2022).

To draw attention to the morphological and molecular features of this rare carcinoma, herein, we describe a series of 6 pulmonary NUT carcinoma cases with different clinical and pathological presentations. Furthermore, we discuss 2 index patients (patients 1 and 2) who exhibited diverse responses to immunotherapy. Patient 1 presented with tumor progression and metastasis when treated with durvalumab concurrent chemoradiotherapy, whereas patient 2 achieved pCR after atezolizumab therapy, suggesting that pulmonary NUT patients might be an excellent beneficiary group for immunotherapy. We also review and discuss the rationale for using immunotherapy to treat pulmonary NUT carcinoma based on published research.

Materials and methods

Patient samples

This study enrolled 6 patients, 3 of whom were identified from the departmental archives in West China Hospital from January 2019 to November 2021, and 3 patients were identified from the consultation files. All cases were reviewed by two senior pathologists (LLJ and WYW). The clinical data, including radiology data, were obtained from the medical records. The pathological information was obtained from pathology reports submitted by the referring pathologists. The follow-up information was collected via medical records plus telephone interviews.

IHC

IHC staining for tumor markers was performed at the time of primary diagnosis according to standard clinical operating procedures. IHC was performed on deparaffinized 4 μm sections with a Ventana Benchmark Ultra automated immunostainer (Ventana Medical Systems, Tucson, AZ, USA). Due to diagnostic uncertainty, a wide panel of primary antibodies was used, including NUT (clone C52B1), PD-L1 (clone 22C3), p63 (clone UMAB4), TTF1 (clone 8G7G3/1), CK5/6 (clone D5/16B4), LCA (clone 2B11), anti-pancytokeratin (clone AE1/AE3), epithelial membrane antigen (EMA) (clone GP1.4), chromogranin A (clone LKZH10), NapsinA (clone MX015), CD56 (clone WAB83), S-100 (clone 4C4.9), desmin (clone MX046), synaptophysin (clone UMAB237), p40 (clone ZR8), CD99 (clone 12E7) and Ki-67 (clone 9-40).

IHC scoring

The percentage of immunoreactivity of each section was semiquantified to evaluate the NUT protein. NUT protein positivity was defined as “strong” when there was speckled nuclear staining in more than 50% of the tumor nuclei, as previously described (Chen et al. 2021). Evaluation of the percentage of tumor cells with partial or complete membranous staining was performed by LLJ and WYW. Positive PD-L1 expression was interpreted as the presence of membranous staining with or without cytoplasmic staining of any intensity in ≥ 1% of tumor cells or immune cells, as described previously (Chen et al. 2021).

FISH

The interphase FISH study was carried out on 5 μm-thick sections generated from FFPE tissues for assessment of the NUTM1 gene rearrangement and BRD4–NUTM1 gene fusion (Anbiping, Guangzhou, China). Briefly, the sections were incubated in a humidified chamber (HYBrite^™ system; Vysis, Abbott, Des Plaines, IL) using two commercially designed probes of bacterial artificial chromosomes flanking the target genes, as previously described. ¹⁷ FISH slides were observed at 100 × magnification under a fluorescence microscope (Leica DM6000, Wetzlar, Germany). Scoring was performed by two independent pathologists (MC and PZ) with expertise in FISH analysis. For the NUTM1 break-apart probe, the cutoff level for scoring as positive was when at least 20% of the nuclei showed a break-apart signal. For the BRD4–NUTM1 fusion probe, the cutoff level for scoring was positive when at least 20% of the nuclei showed fusion signals and several break-apart signals (Chen et al. 2021).

Data analysis

Overall survival (OS) was defined as the time from the start date of the studied treatment to the date of death. OS is reported as the median value expressed in months, with its 95% CI.

Categorical variables were compared using the chi-square test or Fisher’s exact test. Survival analysis was performed using the log-rank test and Kaplan–Meier method. All analyses were performed using SPSS version 19 (SPSS version 19; SPSS, Chicago, IL). P values < 0.05 were considered statistically significant.

Results

Clinical information

The characteristics of the six pulmonary NUT carcinoma patients are summarized in Table 1. There were 5 males and 1 female, with an age at presentation ranging from 31 to 73 years (median 45 years, mean 48 years). The size of the tumors ranged from 1.5 to 13.0 cm in the largest diameter (mean, 6.5 cm). Four of them were located on the right and two on the left, whereas 5 tumors occurred in the lobe and 1 in the trachea. The predominant clinical manifestations were cough (n = 5) and hemoptysis (n = 3), and other symptoms included chest pain (n = 1), back pain (n = 1), dyspnea (n = 1), and fever (n = 1). Only 2 patients (all males) had a history of smoking. All patients were previously healthy, except for patient 4, who had a history of Parkinson’s disease. All patients had advanced-stage disease at the time of diagnosis.

Table 1.

Clinical features of the six patients with pulmonary NUT carcinomas

Patients	Age (years) /sex	Site	Size (cm)	Symptoms	Smoking history	TNM staging	Positive IHC markers	NUTM1 break	BRD4- NUTM1 fusion	Treatment	Recurrence/metastasis	Outcomes/OS
1	45/M	Left lower lobe/trachea	2.5 × 2.5	Cough, headache	Never	pT2bN1M1IVB	NUT, CK5/6, p63, Ki-67 (+ , 60%)	P	P	S + CT durvalumab	Bilateral lung, liver and bone	Dead/4 months
2	45/M	Right lower lobe	longest diameter 9.8 cm	Cough, fever, hemoptysis	Never	cT4N2M0 IIIB	NUT, p63, p53, PCK, CK5/6 (focal), Synaptophysin (focal), Ki-67 (+ , 90%)	P	P	S + CRT atezolizumab	Bilateral lung	Alive/10 months
3	31/M	Left lower lobe	longest diameter 10 cm	Cough, chest pain, hemoptysis	10 + pack-years	cT4N0M0 IIIA	NUT, p63, EMA (focal), Ki-67 (+ , 80%)	P	P	CRT	NA	Dead/4 months
4	73/M	Right lobe	1.3 × 1.5	Cough, hemoptysis, dyspnea	30 + pack years	cT4N0M0 IIIB	NUT, p63, p40, CK7, CK5/6 (focal), CD99 (focal), Ki-67 (+ , 60%), Synaptophysin (focal),	P	P	Support care	NA	Dead/2 months
	49/M	Right trachea	11.2 × 9 × 13	Dyspnea, chest distress, back pain	Never	cT4N2M1c IVB	NUT, p63, p40, CK7, Synaptophysin ( +), Ki-67 (+ , 70%)	P	N	CRT	Bilateral lung	Dead/1.5 months
6	33/M	Right upper lobe	3.4 × 4.5	Cough	Never	cT4N3M0 IIIC	NUT, p63, p40, CK7, Synaptophysin ( +), CD56 ( +), CD34 ( +), Ki-67 (+ , 60%)	P	P	CT	No	Alive/5 months

P positive; N negative; S surgery; CT chemotherapy; Dead died of disease; F female; NA not available; M male; OS overall survival; RT radiotherapy; CRT chemoradiotherapy

Chest plain and enhanced computed tomography (CT) scans revealed irregular low-density soft shadows in all six patients (Fig. 1A). The lesions were large, with the longest diameter ranging from 1.5 to 13 cm (Fig. 2A1 and A2). Two patients (2 and 5) had unresectable masses: one was 9.8 cm, and the other was 13 cm. Four of the 6 lesions were centrally located. Lesions invaded the ipsilateral lung and fused with the ipsilateral and mediastinal lymph nodes. Three patients (2, 4, and 5) had small or limited pleural effusion, and 2 patients (2 and 5) had obstructive pneumonia. Patient 5 had primary tracheal lesions and presented with disease in the thoracic cavity and lungs. At the time of diagnosis, 50% (3/6) of the patients showed regional infiltrating lymph nodes, and no patient presented with distant metastases.

Fig. 1.

Imaging characteristics of patient 1. A Contrast-enhanced CT scan of the chest demonstrated a mass in the left upper lobe and was a central type with moderate to large pleural effusion on the left (black arrow). B Contrast-enhanced CT showed liver lesions (black arrow). C A whole-body FDG-PET image demonstrated marked FDG avidity of the dominant lung lesion, measuring 0.3–0.8 cm, as well as metastatic sites, including innumerable osseous lesions and a subcutaneous lesion in the right flank. Bone metastasis was present, including lytic lesions (black arrow)

Fig. 2.

Imaging characteristics of patient 2. (A1, A2) Lesions were observed before treatment in the CT images in June 2021. (B1, B2) The lesions shrunk to 5.3 × 2.2 cm after 3 cycles of chemoradiotherapy concurrent with atezolizumab in July 2021. (C1, C2) After another three cycles of treatment, the mass lesion size continued to decrease to 3.1 × 2.6 cm in November 2021. (D1, D2) The mass lesion size continued to decrease to 2.8 × 2.2 cm in December 2021

Histological features

The neoplastic cells grew infiltratively and were closely associated with the bronchial mucosa membrane (Fig. 3A). Occasionally, alveolar epithelial hyperplasia was noted (Fig. 3B). Typically, the tumor cells were composed of poorly differentiated neoplasms that predominantly grew in poorly cohesive sheets or nests, and abrupt keratinization was found in 50% (3/6) of patients (Fig. 3C). In addition, the tumor cells shared some common morphological features, such as high nuclear-to-cytoplasmic ratios, round to oval nuclei with irregular contours that were uniform in size, and areas of coagulative necrosis. Surprisingly, in patient 2, no cancer cells (Fig. 3E), but only fibroreactive and inflammatory cells with hemosiderin (Fig. 3F), could be identified in the resected pulmonary tissue and lymph nodes under postoperative pathological examination, confirming pCR in the lung tumor after treatment.

Fig. 3.

Pathological features of pulmonary NUT carcinoma. A The neoplastic cells showed an infiltrative growth pattern and were closely associated with the bronchial mucosa membrane. B Alveolar epithelial hyperplasia. C Typically, the tumor cells were composed of poorly differentiated neoplasms predominantly growing in poorly cohesive sheets or nests with focal abrupt keratinization. D Cytologically, a few cells showed overt pearl formation, suggesting keratinization. E No cancer cells could be identified in the resected pulmonary and lymph nodes, indicating a pathologic complete response in patient 2. F Fibroreactive and inflammatory cells with hemosiderin in the removed lobe of patient 2

Cytological materials were available for patients 2 and 4. Papanicolaou-stained smear showing closely cohesive clusters. Both of the specimens displayed evidence of squamous differentiation, such as large epithelial cells with dense cytoplasm, and a few cells showed overt pearl formation, suggesting keratinization (Fig. 3D). Cytoplasmic vacuoles are visible, with high nuclear/cytoplasmic ratios, polymorphic nuclei, slightly irregular nuclear contours, and identifiable mitotic figures.

IHC studies

The IHC results are summarized in Table 1 and showed consistent immunophenotypes. Six patients (6/6) were positive for NUT, including speckled nuclear staining in five patients (Fig. 4A) and focal staining in one patient. The squamous cell lineage exhibited p63 (6/6), CK5/6 (2/3), CK7 (3/4), and pancytokeratin (1/2) expression, ranging from focal to diffuse nuclear staining. Four of these 5 patients were positive for synaptophysin, with one being diffuse and the other being partial (Fig. 4B). Additional focal or weak staining of CD34 (1/1), EMA (1/1), CD99 (1/1), chromogranin A (1/4), CD56 (1/3), and p53 (1/1) was noted in some patients. Furthermore, all six patients had negative PD-L1 expression, except patient 1, who had focal immune cell staining, which was less than 1%. Other markers, including S100, desmin, TTF1, LCA, and Napsin A, all showed negative results. The Ki-67 index ranged from 60 to 90% (mean, 70%).

Fig. 4.

Immunohistochemical and molecular features of pulmonary NUT carcinoma. A Diffuse nuclear staining for NUT. B Focal positivity for synaptophysin. C FISH analysis shows NUTM1 gene rearrangement in all 6 patients. D FISH analysis shows BRD4–NUTM1 fusion in patients 1-4 and 6

FISH results

All six patients showed NUTM1 gene 15q14 rearrangement (Fig. 4C), which was present in 88, 85, 88, 85, 75 and 62% of the tumor cells, respectively. Meanwhile, five of the six patients (patients 1–4, 6) showed BRD4–NUTM1 fusion, whereas patient 5 was negative for BRD4–NUTM1 fusion (Fig. 4D), with an unknown NUTM1 variant. Further next-generation sequencing was performed for patient 5. Unfortunately, the result was unsatisfactory because of insufficient DNA.

Treatment and outcomes

The treatment and outcomes of the six pulmonary NUT carcinomas are summarized in Table 1. The median overall survival (OS) was 2.9 months (1.5–4 months). The initial treatments included chemotherapy (4/6, 66.7%) and surgery (1/6, 14.3%). Two patients (1 and 2) received ICIs (programmed cell death ligand 1 [PD-L1] monoclonal antibody) as second- or higher-line treatments and are discussed further below due to their diverse responses relative to an absence of PD-L1 expression. Patient 4, a 73-year-old man, was too old to receive standard systemic surgery or chemoradiation and was observed to be off-treatment. He died of the disease 2 months later as a result of respiratory failure. Patient 5 received only one cycle of chemotherapy (albumin-bound paclitaxel 400 mg/m² d1, carboplatin 40 mg/m² d1, d2, 30 mg/m² d3 and bevacizumab every 3 weeks) and radiotherapy from November 2021 to December 2021. He quickly experienced tumor recurrence and metastasis to the bilateral lung and died of the disease, having the shortest survival time (1.5 months).

Patient 1 underwent surgical resection one month later in September 2020. The patient received three cycles of TP chemotherapy concurrent with durvalumab (albumin-bound paclitaxel 400 mg/m² d1, cisplatin 60 mg/m² d1, d2, 30 mg/m² d3 and durvalumab 1000 mg d1 every 3 weeks) from October 2020 to January 2021. Unfortunately, even with high-intensity management, this patient experienced tumor recurrence and metastasis to the liver and bone (Fig. 1B and C) and died of the disease 4 months later.

Patient 2 received preoperative EP chemotherapy (etoposide 150 mg/m² d1–d3, cisplatin 40 mg/m² d1–d3, combined with atezolizumab 1200 mg/m² d1) and radiotherapy every three weeks. After three cycles of treatment, the mass in the right lower lobe shrank to 5.3 × 2.2 cm (Fig. 2B1 and B2). After another three cycles of treatment, the mass lesion size continued to decrease to 3.1 × 2.6 cm in November 2021 (Fig. 2C1 and C2). The mass lesion size decreased even further to 2.8 × 2.2 cm in December 2021 (Fig. 2D1 and D2). The tumor mass was relatively small; therefore, right upper lobectomy combined with lymph node dissection was performed in March 2022. Complete response (CR) was achieved based on CT evaluation without evidence of recurrence or metastasis after 10 months.

Review of studies

We reviewed previously reported pulmonary NUT carcinomas and identified 15 patients who received ICIs (including two cases in this study) (Table 2). These patients demonstrated a significantly prolonged median survival time compared to patients who did not receive ICIs (13.8 vs 4.4 months), ranging from 2.2 to 79 months. In addition, three out of 15 patients presented a complete response to the immunotherapy and concurrent chemotherapy regimen. PD-1 inhibitors may be better than PD-L1 inhibitors in patients with pulmonary NUT carcinoma (Fig. 5, p = 0.061).

Table 2.

Clinicopathological features, immunotherapy, and outcomes

Patients	Age (years) /sex	PD-L1 IHC TPS (%)	TMB (muts/Mb)	NUTM1 fusion	Treatment (PD-1/PD-L1 inhibitor)	Response to immunotherapy	OS (months)	References
1	31/F	NA	1.75	CHRM5-NUTM1	DP + cetuximab (nivolumab + pembrolizumab)	NA	12+	Xie et al. (2020)
2	34/M	80%	NA	NA	CT (pembrolizumab)	NA	12+	Cho et al. (2020)
3	45/M	70%	NA	NA	RT + CT (pembrolizumab)	NA	12+
4	48/F	0%	NA	NA	CT (pembrolizumab)	NA	12+
5	31/M	NA	NA	NSD3-NUTM1	RT + CT (atezolizumab)	NA	2.2	Hung et al. (2021)
6	53/M	NA	NA	BRD3-NUTM1	RT + CT + HDAC inhibitor (nivolumab)	NA	11.6
7	57/M	NA	NA	NA	Cisplatin + docetaxel (nivolumab)	Disease progression	4	Maruyama et al. (2018)
8	31/M	10%	NA	NA	S + adjuvant CT (nivolumab)	Complete response	79+	Davis et al. (2021)
9	34/F	0%	NA	NA	CT (pembrolizumab)	NA	Lost	Joel et al. (2020)
10	52/F	0%	0.8	BRD4–NUTM1	EP + irinotecan and docetaxel (nivolumab)	Complete response	4	Riess et al. (2021)
11	39/F	1%	1	BRD4–NUTM1	Chest RT + EP (atezolizumab)	Disease progression	5
12	49/M	NA	NA	BRD4–NUTM1	S + RT + CT (durvalumab + pembrolizumab)	Partial response	18	Gupta et al. (2022)
13	38/M	30%	NA	BRD4–NUTM1	CT (camrelizumab)	NA	10	Zhou et al. (2022)
14	45/M	0%	NA	BRD4–NUTM1	S + TP (durvalumab)	Disease progression	4	This study
15	45/M	0%	NA	BRD4–NUTM1	EP + RT (atezolizumab)	Partial response	Alive, 8+

M male; F female; S surgery; CT chemotherapy; RT radiotherapy; OS overall survival; IHC immunohistochemistry; TMB tumor mutation burden; PD-L1 programmed cell death ligand 1; NA not available; EP etoposide + cisplatin; TP albumin-bound paclitaxel + cisplatin

Fig. 5.

Kaplan–Meier curves for overall survival (OS) in NUT carcinoma patients treated with PD-1 inhibitors (nivolumab, pembrolizumab and camrelizumab) and two PD-L1 inhibitors (atezolizumab and durvalumab). P = 0.061

Discussion

Pulmonary NUT carcinoma has been classified by the WHO as a novel type of lung cancer [1]. It is extremely rare and often difficult to distinguish from other poorly differentiated tumors. In our single center, a total of 12 pulmonary NUT carcinomas were identified, and six cases have been previously published (Chen et al. 2021). Of the 12 cases included, the age at presentation ranged from 22 to 73 years (median 43 years, mean 43.4 years). In addition, ten patients were male, and two were female, showing a predilection toward males. Notably, pulmonary NUT carcinoma is lethal, regardless of whether intensive resection is performed followed by adjuvant chemotherapy or radiation. Thus, in this study, we focused on ICIs to explore the rationality of such therapy based on published research. Previous reports on immunotherapy for pulmonary NUT carcinoma were also briefly reviewed.

NUT carcinoma can be diagnosed through a combination of CT findings, histological features, NUT IHC staining, and molecular tests, such as FISH and next-generation sequencing. IHC staining showed consistent immune profiles, and all patients showed NUT positivity with variable expression of the squamous cell lineage, including p63, CK, and p40 expression. Some reports have observed patchy staining of synaptophysin and TTF1 (Hung et al. 2021; Pezzuto et al. 2020). It is worth noting that there were 4 patients in this case series who showed immunoreactivity for synaptophysin. The case reported by Hung et al. was also diffusely immunoreactive for neuron-specific enolase (2021). Together with CD56 staining, the aberrant expression of neuroendocrine markers in a few cases may lead to a misdiagnosis of neuroendocrine tumor.

Genetically, BRD4 is the most common fusion partner of NUTM1, accounting for approximately 78% of rearrangements. Other fusion partners, including BRD3 and NSD3, ZNF532 and ZNF592, have also been reported (Chau et al. 2019). The BRD4 protein can colocalize and interact with BRD3, NSD3, ZNF532, ZNF592, ZNF687, and ZMYND8 to form chromatin regulatory complexes. In our cohort of 12 patients with pulmonary NUT carcinoma, nine (9/12, 75%) patients had the BRD4–NUTM1 fusion. Two patients harbored the NSD3-NUTM1 variant (2/12, 17%). However, the remaining patient had no detailed genetic data. Recently, new fusion partners for NUTM1, such as CIC, BCORL1, MGA, and MXD4, have been identified in NUTM1-rearranged neoplasms, including primary undifferentiated soft tissue tumors, visceral tumors, sarcomas, and epithelioid malignancies (Chen et al. 2022).

Lee et al.’s (2017) findings suggest that a single catastrophic event in proliferating normal cells could be sufficient for neoplastic transformation into NUT carcinoma. Consistent with this study, several studies have focused on the genomic profiling of NUT carcinoma through whole-genome or whole-exome sequencing, showing that NUT carcinoma rarely has mutations in tumor suppressor genes or classic canonical oncogenes (Cavalieri et al. 2017; Xie et al. 2021; Zhang et al. 2021). Moreover, sequencing analysis showed that pulmonary NUT carcinoma has a lower mutation rate than other types of NSCLC. Furthermore, MYC plays a key role in the pathogenesis of NUT carcinoma. Previous studies demonstrated that MYC is a downstream target of BRD-NUT, and targeting MYC was necessary and sufficient for the blockade of NUT carcinoma differentiation (Grayson et al. 2014). In our previous study, diffuse MYC expression, trisomy 8 and three copies of the MYC gene were noted in two parotid gland NUT carcinomas (Chen et al. 2022). The MYC/chromosome 8 amplification probe displayed three red signals and three green signals in most tumor cells. The results indicated MYC upregulation in NUT carcinoma.

ICIs have been used as novel cancer therapies. In particular, a number of clinical trials exploring the use of anti-PD-1/anti-PD-L1 monoclonal antibodies (mAbs) in various types of cancer are ongoing worldwide. PD-L1 expression has been regarded as a predictive biomarker of the efficacy of ICIs. PD-L1 expression assessment prior to the administration of treatment is important for identifying patients who are likely to benefit from ICIs. Several antibodies, including clone 22C3 and clone SP142, have been approved as a companion with ICI administration by the Food and Drug Administration (FDA). However, very little information is available on PD-L1 staining in NUT carcinoma. Recently, several studies have reported high PD-L1 expression in NUT carcinomas (Cho et al. 2020; Gupta et al. 2022; Zhang et al. 2021; Pan et al. 2020). Cho et al. reported the highest PD-L1 expression in 2 out of 10 cases (70% and 80%, respectively). In He et al.’s study (2020), they found that PD-L1 ranged from 20% TPS in an adult case to 0 or 1% in pediatric cases. In contrast, another study including 13 NUT patients reported by Jung M et al. indicated that only one case had 10% PD-L1 expression (Jung et al. 2019). Similarly, in our present and previous studies, none of the 12 pulmonary NUT carcinomas showed PD-L1 staining. From these limited studies, we conclude that 13.5% (5/37) of NUT carcinomas display PD-L1 expression, although it is difficult to accurately evaluate PD-L1 expression in NUT carcinoma due to its rarity.

To date, the immunotherapy drugs mentioned in the reported literature include three PD-1 inhibitors (nivolumab, pembrolizumab and camrelizumab) and two PD-L1 inhibitors (atezolizumab and durvalumab) (Li et al. 2021). Recently, nivolumab and pembrolizumab received accelerated approval by the US FDA for the treatment of extensive-stage small-cell lung cancer (ES-SCLC) (Antonia et al. 2017). Meanwhile, atezolizumab in combination with CP and ET (CP/ET) and durvalumab plus platinum and ET were also approved for the first-line (1-L) treatment of ES-SCLC by the US FDA (Elegbede et al. 2021; Lee et al. 2021). Exploratory analyses suggest that ICIs conferred clinical benefit to these patients. However, the efficacy of such combination immunotherapy in patients with pulmonary NUT carcinoma is unknown.

In this cohort, we presented two index pulmonary NUT cases, both showing negative PD-L1 expression but with diverse responses to ICIs with concurrent chemotherapy. Patient 1 underwent durvalumab without any success. This is consistent with the report by Gupta et al. who first prescribed consolidation therapy in an NUT patient with durvalumab concurrent chemoradiation. This patient developed metastases within 2 months of treatment. It seems that durvalumab does not improve the prognosis of advanced NUT patients. Patient 2, who had an unresectable mass, was treated successfully with atezolizumab, showed stable disease 6 months after therapy initiation and experienced a marked improvement in symptoms. This finding is consistent with the notion that some patients with high PD-L1 expression do not respond to ICIs, whereas a small proportion of patients with no PD-L1 expression do respond to ICIs.

The efficacy of combination immunotherapy may be related to other tumor-related markers, such as the tumor mutational burden (TMB), microsatellite instability (MSI) and mismatch repair deficiency (dMMR). Generally, patients with high PD-L1 expression, a high TMB, and MSI are more likely to benefit from ICIs. We expect the same to be true for patients with NUT carcinoma. We reviewed previously reported pulmonary NUT carcinomas and identified 15 patients who received ICIs (including two cases in this study). These patients demonstrated a significantly prolonged median survival compared to patients who did not receive ICIs, and PD-1 inhibitors may be better than PD-L1 inhibitors in patients with pulmonary NUT carcinoma. These findings suggest that ICIs need to be studied further in this patient population.

This study has some limitations. First, this was a retrospective study from a single center. Second, because of the rarity of pulmonary NUT carcinoma cases, the number of samples was small. Finally, because of insufficient tumor samples, we did not investigate the tumor-immune microenvironment in pulmonary NUT carcinoma patients. Therefore, there may be bias in the analysis of the clinicopathological features and prognosis of pulmonary NUT carcinoma patients. We will screen more samples for further research in the future.

Conclusions

The present findings prove that pulmonary NUT carcinoma exhibits squamous differentiation similar to other NUT carcinomas. It displays diffuse NUT expression and NUTM1 rearrangement, occasionally with focal synaptophysin staining. These results indicate that it is essential to detect NUT protein expression and NUTM1 rearrangement to differentiate NUT carcinoma from other poorly differentiated carcinomas. Despite intensive treatment, the overall survival time was only 2.9 months, and these results also emphasize the importance of exploring new therapies for pulmonary NUT carcinoma. Without the expression of PD-L1, a diverse response to immunotherapy was observed in this study. Furthermore, a patient responded remarkably to ICIs and achieved pCR without uncontrollable adverse events. This is the first NUT carcinoma with pCR. However, given the limited number of cases, further research needs to be conducted to investigate the tumor-immune microenvironment and immunotherapy in NUT carcinoma in more depth.

Author contributions

MC collected the patients, performed the experiments, and wrote the paper; YZ performed the immunohistochemistry experiments; XC and WW performed the histopathological examinations; and LJ organized and modified the paper. All authors were involved in analyzing the study data and commenting on the manuscript. All authors approved the final manuscript.

Funding

This study was supported by the Sichuan University 1.3.5 Project for Disciplines of Excellence-Clinical Research Incubation (No. 2019HXFH002).

Availability of data and materials

The original contributions presented in the study are included in the article/supplementary materials. Further inquiries can be directed to the corresponding author.

Declarations

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.