SMARCB1/INI-1-Deficient sinonasal carcinoma demonstrates a poor prognosis but favorable clinical outcomes after PD-1/PD-L1 inhibitor therapy: A case series

Abstract

Despite advances in multimodal cancer therapy, such as combining radical surgery with high-intensity chemoradiotherapy, for SMARCB1/INI-1-deficient sinonasal carcinoma (SDSC), the prognosis of patients remains poor. Immunotherapy is gaining increasing popularity as a novel treatment strategy for patients with SMARCB1/INI-1-deficient tumors. Herein, we report on the management of three patients with SDSC who received PD-1/PD-L1 inhibitor therapy as a part of multimodal therapy based on surgery and chemoradiotherapy. All three patients survived and demonstrated good clinical remission and disease control. To our knowledge, this is the first case series reporting the use of immunotherapy to improve clinical outcomes in neoadjuvant, adjuvant, and late first-line stages of treatment in patients with SDSC. Furthermore, we reviewed the relevant literature and further explored the correlation between SMARCB1/INI-1 deletion and immunotherapy.

Introduction

SWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily B member 1/integrase interactor-1 (SMARCB1/INI-1), one of the core subunit proteins encoding the adenosine triphosphate-dependent SWI/SNF chromatin remodeling complex, is a tumor suppressor gene located on chromosome 22q11.2.^1,2 SMARCB1/INI-1 deletion was initially recognized as a distinct feature of malignant rhabdoid tumor (MRT) and is also associated with the onset and progression of epithelioid sarcoma (ES), poorly differentiated chordoma (PDC), renal medullary carcinoma (RMC), rhabdoid carcinoma of the pancreas or gastrointestinal tract, and sinonasal carcinoma.^1–3 In 2014, SMARCB1/INI-1-deficient sinonasal carcinoma (SDSC) was first reported by Agaimy et al. and Bishop et al. and accounts for only approximately 1% of all head and neck malignancies.^4–6 In 2017, the World Health Organization recognized SDSC as a new subtype of sinonasal undifferentiated carcinoma (SNUC) pathologically characterized by variable rhabdoid cell components and complete deletion of SMARCB1/INI-1.^4,5,7,8

Owing to the rarity of SDSC, the treatment strategies used currently are based on the treatment approaches followed for other types of sinonasal carcinomas. A multidisciplinary approach combining surgery with chemoradiotherapy is the accepted treatment strategy.^6,8,9 At the time of initial diagnosis, approximately 80% of patients with SDSC have T4 stage disease, making radical surgeries challenging because of extensive tumor invasion. ⁶ In recent years, neoadjuvant therapy has emerged as a promising therapeutic modality in patients with SDSC, as it can either downstage the tumor for radical surgery or be combined with concurrent chemoradiotherapy to achieve results comparable to radical surgery.^10–12 Despite the availability of multimodal therapies, the prognosis of patients with SDSC remains poor with a 3-year overall survival (OS) rate of 51.8%, ⁶ necessitating the development of novel treatment strategies.

The immunogenicity of SMARCB1/INI-1-deficient tumors has been explored and validated in recent years. Among 30 pediatric patients with sarcoma who had negative SMARCB1/INI-1 expression, 47% showed positive programmed death-ligand 1 (PD-L1) expression. ¹³ The use of ICIs, including programmed cell death protein 1 (PD-1)/ PD-L1 inhibitors, as mono- or combination therapy for SMARCB1/INI-1-deficient tumors has also achieved clinical benefits.^14–17 Meanwhile, ICIs are gaining increasing recognition for the treatment of head and neck cancer, and clinical trials are recruiting patients with SMARCB1/INI-1-deficient tumors for ICI therapy.^6,18 Therefore, immunotherapy is emerging as one of the promising oncotherapeutic strategies available for SDSC.

In this study, the treatment and clinical outcomes of three patients with SDSC are reported. PD-1/PD-L1 inhibitors were used in neoadjuvant, adjuvant, or late first-line stages of treatment. All three patients are alive and have achieved good disease control and remission. To our knowledge, this is the first case series of the early application of immunotherapy in the management of patients with SDSC. In addition, the relevant literature was reviewed and the correlation between SMARCB1/INI-1 deletion and immunotherapy was further explored.

Case reports

Case 1

A male adult patient presented to our hospital in August 2021 with a chief concern of “nosebleeds for more than 3 months.” Outpatient nasal mass biopsy revealed SDSC with immunohistochemical staining showing negativity for INI-1. Enhanced MRI showed a nodular soft tissue signaling shadow measuring approximately 0.7 × 1 × 1.9 cm in the right frontal sinus and anterior ethmoid sinus (Figure S1d and S1e). The patient subsequently underwent radical surgical resection, achieving R0 resection with negative surgical margins. Postoperative MR reexamination findings are shown in Figure S1f and S1g. A final diagnosis of SDSC pT3N0M0 was established based on the staging examination results and postoperative pathologic findings (Figure S1a–S1c). Immunohistochemical staining results showed partial positivity for pan-cytokeratin (PCK), epithelial membrane antigen (EMA), P40, P63, and CD99; focal positivity for cytokeratin (CK) 5/6 and synaptophysin (Syn); and negativity for INI-1, P16, CD56, CgA, SOX10, S-100, nuclear protein in testis (NUT), CD34, and desmin. The Ki67 proliferation index was approximately 80%, and the PD-L1 Combined Positive Score (CPS) was <1.

Postoperatively, the patient first received adjuvant chemotherapy with the EP (etoposide + cisplatin) regimen and sequential radiotherapy. During radiotherapy, the EP regimen combined with the PD-L1 inhibitor durvalumab was administered for two cycles of synchronized treatment. After radiotherapy, the patient received durvalumab every 3 weeks with regular follow-up. As of the last follow-up on October 12, 2024, the patient was still alive and showed no signs of recurrence or progression (Figure S1h–S1i), with a disease-free survival (DFS) of 38 months.

Case 2

A female adult patient presented to our hospital in February 2022 with a chief concern of “nosebleeds for more than 1 month.” A nasal mass biopsy revealed SDSC (Figure 1(a)–(c)). Immunohistochemical staining results showed positivity for PCK, EMA, CK7, CK8/18, CDX2, and BRG-1; partial positivity for P63, P40, thyroid transcription factor-1, and desmin; and negativity for INI-1, CK20, CK5/6, PAX8, S100, leukocyte common antigen, Syn, NUT, SOX10, and GATA-3. The Ki67 proliferation index was approximately 40%, and the CPS was 10. PET/CT showed a shadow measuring 4.2 × 3.2 cm in the nasal cavity and sinuses, with no metastatic signs. Therefore, the clinical staging of the tumor was determined as cT4N0M0.

Figure 1.

Pathology and imaging information for case 2 with timeline. (a) The tumor tissue is arranged in solid sheets and nests, displaying basaloid morphology (HE staining with magnification ×40). (b) The basaloid tumor cells have scant cytoplasm, resulting in a high nuclear-to-cytoplasmic ratio. The nuclei exhibit vacuolation and prominent nucleoli, with frequent mitotic figures observed (HE staining with magnification ×100). (c) The tumor cells show a loss of SMARCB1/INI-1 expression, with nuclear positivity in stromal cells serving as an internal control (immunohistochemical staining with magnification ×200). (d) Primary lesions on MRI T1-weighted image at recurrence (arrow); (e) primary lesions on MRI T2-weighted image at recurrence (arrow); (f, g) pulmonary metastases on CT at recurrence (arrow); (h) primary lesions reaching PR on MRI T1-weighted image after 2 cycles of first-line treatment (arrow); (i) primary lesions reaching PR on MRI T2-weighted image after 2 cycles of first-line treatment (arrow); (j, k) Pulmonary metastases reaching PR on CT after 2 cycles of first-line treatment.

Considering the late local stage of the tumor and the difficulty of surgical resection, the patient was treated with neoadjuvant EP chemotherapy combined with the PD-1 inhibitor toripalimab. After receiving one treatment cycle, the patient returned to the local hospital because of the outbreak of the COVID-19 epidemic where the patient received the chemotherapy combined with toripalimab for two cycles and sequential nasal radiotherapy. Subsequently, the patient underwent radical surgical resection, but microscopic residual disease (R1 resection) was identified. The patient then received adjuvant chemotherapy with EP regimen and oral S-1 maintenance therapy.

In October 2023, the patient presented with “distension and blurred vision in the right eye.” MR reexamination showed a mixed signal shadow in the nasal cavity measuring 52 × 38 mm (Figure 1(d) and (e)). Lung CT showed multiple solid nodules in both lungs (Figure 1(f) and (g)), suggesting tumor recurrence and metastasis. Therefore, the patient received a TP (paclitaxel + platinum) regimen of chemotherapy combined with toripalimab as advanced first-line treatment. MR performed after two cycles of treatment showed a significant reduction in intranasal mass (Figure 1(h) and (i)) and pulmonary metastasis (Figure 1(j) and (k)), and the efficacy of immunotherapy was evaluated as partial response (PR). As of the follow-up on January 11, 2024, the patient was still alive and showed no signs of tumor progression. Unfortunately, due to financial constraints, the patient discontinued treatment, and on May 15, 2024, a reexamination at the local hospital indicated disease progression, with an OS of 27 months.

Case 3

A female adult patient was admitted to our hospital in March 2023 with a chief concern of “nosebleed and headache.” A sinus MR was performed in another hospital that revealed bilateral sphenoid sinus and left ethmoid sinus tumors. Thus, the patient underwent residual surgical resection, with macroscopic residual disease, resulting in R2 resection. The postoperative histopathological findings suggested SDSC (Figure S2a and S2b). Immunohistochemical staining showed positivity for CKP, CK8, BRG-1, and SOX2; partial positivity for P63, Syn, and CD99; and negativity for INI-1, CD56, CgA, S100, SOX10, calretinin, NF, glial fibrillary acidic protein (GFAP), NKX2.2, PAX5, thyroid transcription factor-1, Bcl2, and vimentin. The Ki67 proliferation index was approximately 70%. Based on these findings and stage examination, a definitive diagnosis of SDSC pT4N0M0 was established.

Postoperative MR reexamination suggested a residual lesion at the base of the sphenoid sinus (Figure S2c and S2d). Therefore, the patient received paclitaxel combined with the PD-1 inhibitor pembrolizumab. MR reexamination findings after two cycles of treatment suggested that the lesion had reduced in size, and the effectiveness of immunotherapy was determined as PR (Figure S2e and S2f). The patient received sinus concurrent chemoradiotherapy four months postoperatively. After radiotherapy, the efficacy of immunotherapy was determined as complete response (CR) on imaging (Figure S2g and S2h). The patient received oral capecitabine for maintenance therapy and immunotherapy with pembrolizumab every 3 weeks. As of the last follow-up on October 11, 2024, the patient was alive and showed no signs of recurrence or progression, with a DFS of 19 months.

The reporting of this study conforms to CARE guidelines. ¹⁹ We have de-identified all patient details to ensure anonymity and obtained the patient's consent for treatment.

Discussion

SDSC accounts for only 3% to 6% of all sinonasal carcinomas, with only about 200 cases reported worldwide so far.^6,20 Therefore, there is a lack of large-sample clinical studies exploring standard treatment strategies for SDSC. Although the optimal sequence of implementation is not yet clear, the integrated treatment model of surgery combined with chemoradiotherapy is the most recognized treatment strategy at this stage. Unfortunately, SMARCB1/INI-1 deficiency is associated with a worse prognosis compared to SNUC.^3,6,15,21 An objective response rate (ORR) of only 30% has been observed in patients with SDSC receiving neoadjuvant chemotherapy. ¹² Several systematic reviews have shown a median survival of only 22 months and a 3-year OS rate of only 51.8% in patients with SDSC.^6,22,23 Table 1 summarizes the characteristics, treatment history, and clinical outcomes of our three patients. Similar to previous studies, we combined surgery with chemoradiotherapy to manage our patients. ⁸ However, we added PD-1/PD-L1 inhibitors to each of the neoadjuvant, adjuvant, or late first-line treatment phases. All our patients survived, and two of them showed no obvious signs of tumor recurrence or progression. Despite disease recurrence and metastasis in one patient due to treatment interruption, our late first-line treatment resulted in excellent disease control. With the same application of surgery combined with chemoradiotherapy, a more favorable clinical outcome was observed, possibly indicating the critical role of immunotherapy.

Table 1.

Patient characteristics and clinical outcomes.

	Case 1	Case 2	Case 3
Site	Frontal sinus and ethmoid sinus	Nasal cavity, maxillary sinus, sphenoid sinus, and frontal sinus	Sphenoid sinus and ethmoid sinus
Stage	pT3N0M0	cT4N0M0	pT4N0M0
Neoadjuvant therapy	None	CT + anti-PD-1	None
Surgery	Surgery without residuals	Surgery with residuals	Surgery with residuals
Adjuvant therapy	CT + CCRT + anti-PD-L1	CT + RT	CT + anti-PD-1 + CCRT
Maintenance therapy	Anti-PD-L1	Oral CT	Oral CT + anti-PD-1
Progression	None	Recurrence and metastasis	None
First-line treatment	None	CT + anti-PD-1	None
Best response	NA	PR	CR
Survival (mo)	DFS (>38 mo)	OS (>27 mo)	DFS (>19 mo)

CT: chemotherapy; RT: radiotherapy; CCRT: concurrent chemoradiotherapy; PD-1: programmed cell death protein 1; PD-L1: programmed death-ligand 1; NA: not applicable; PR: partial response; SD: stable disease; mo: months; DFS: disease-free survival; OS: overall survival.

Microscopically, SDSC exhibits rhabdomyoid-like histological features, often observed with tumor necrosis and a high mitotic rate.^7,24 Immunohistochemistry shows variable positivity for squamous and neuroendocrine markers, and negativity for SMARCB1/INI-1 and NUT.^7,24 The present Case 2 exhibited positive CDX2 expression, which is rare in SDSC. While CDX2 positivity is common in colorectal adenocarcinoma and intestinal-type sinonasal adenocarcinoma (ITAC), it has also been reported in SDSC with yolk sac tumor differentiation and SNUC.^25–28 This highlights the complexity of sinonasal tumors, suggesting CDX2 may not be the only marker for intestinal differentiation. Additionally, TTF-1 positivity, commonly observed in tumors of thyroid follicular origin, pulmonary adenocarcinoma, and small cell lung cancer, and Desmin, a muscle-specific marker typically expressed in muscle-derived tumors, were tested in our cases.^29,30 Case 2 showed partial positivity for both markers, while Cases 1 and 3 were negative. However, the positive expression of TTF-1 and Desmin in SDSC has not been reported, underscoring the need for further research into potential immunophenotypic variations in SDSC.

As a member of the SWI/SNF complex, SMARCB1/INI-1 has an immunomodulatory role in which the potential link between SMARCB1/INI-1 deletion and ICI therapy has been recently explored. ³¹ Leruste et al. demonstrated in MRT that SMARCB1/INI-1 deficiency mediates tumor immunogenicity through the desuppressive function of multiple endogenous retroviral elements. ³² Activation of the cGAS/STING pathway by SMARCB1 /INI-1 deletion in RMC is also thought to correlate with enhanced tumor immunogenicity. ³³ Moreover, higher PD-L1 expression rates were detected in tumors with negative SMARCB1/INI-1 expression.^13,34,35 Kim et al. also confirmed high PD-L1 expression in all seven patients with ES. ³⁶ In our case series, Case 2 had a CPS of 10, consistent with the trend of high PD-L1 expression reported in the literature, which may explain the positive response to immunotherapy. However, despite a CPS of less than 1, Case 1 also demonstrated good disease control following immunotherapy, suggesting that immunotherapy efficacy in SDSC may not rely solely on high PD-L1 expression. This finding aligns with the results of the CHECKMATE-141 trial, which similarly failed to show a clear link between PD-L1 expression and immunotherapy outcomes in head and neck squamous cell carcinoma, highlighting the controversy and limitations of PD-L1 as a biomarker. ³⁷ Due to the unavailability of the specimen, PD-L1 expression for Case 3 could not be obtained. Future research with larger cohorts is needed to better understand the relationship between SMARCB1/INI-1 deficiency, PD-L1 expression, and the effectiveness of ICI therapy. On the basis of the potential association between the SMARCB1/INI-1 gene and immunotherapy, the use of ICIs in the treatment of SMARCB1/INI-1-deficient tumors appears to be reasonable and acceptable.

In recent years, PD-1/PD-L1 inhibitors have been used for treating various types of SMARCB1/INI-1-deficient tumors. Table S1 summarizes the diagnostic and therapeutic information and clinical benefits of the relevant studies to date.^13,38–54 A patient with ES and PD-L1-positivity achieved PR after receiving pembrolizumab monotherapy in the KEYNOTE-051 study. ⁴² Patients with MRT treated with the PD-L1 inhibitor atezolizumab also demonstrated transient objective response.^43,55 In tumors of non-mesenchymal tissue origin such as metastatic pancreatic cancer and non-small-cell lung cancer, patients with SMARCB1/INI-1 gene alterations respond well to ICI therapy.^39,40 However, the effectiveness of immunotherapy in SDSC remains to be elucidated. PD-1/PD-L1 inhibitors were added to the combined therapy in all patients with SDSC, which provided favorable clinical outcomes, justifying its efficacy in multimodal therapy by improving patient prognosis in this report.

Except for classical Hodgkin's lymphoma, the efficacy of ICI therapy in pediatric tumors has been unsatisfactory. However, pediatric SMARCB1/INI-1-deficient tumors demonstrates increased susceptibility to ICIs. ⁵⁶ Similarly, the use of immunotherapy in other SNUCs remains to be elucidated.^57,58 In the present case series, favorable patient response to PD-1/PD-L1 inhibitors was observed, which suggests an association between SMARCB1/INI-1 deficiency and susceptibility to immunotherapy. Although studies included only small numbers of patients, mutations in the SWI/SNF complex genes may be predictive markers of ICI therapy efficacy.^15,16,59,60 Bakouny and Fountzilas et al. further suggested that the deletion of SMARCB1/INI-1 may be associated with immunotherapy positive response in patients with MRT.^61,62 Based on the conclusion that our patients achieved good disease control with immunotherapy, the predictive ability of SMARCB1/INI-1 deletion on the efficacy of ICI therapy may be extendable to other SMARCB1/INI-1-deficient tumors, including SNUC.^61,62 However, similar to indicators such as PD-L1 expression, microsatellite instability status, and tumor mutational burden, progressive disease resulting from ICI therapy suggests that SMARCB1/INI-1 is unlikely to be a perfect predictive marker.^{38,41–43,45,46,51,63} Therefore, future large-sample-based predictive modeling is required to explore its predictive ability.

Herein, we report on the management of three patients with SDSC. They have survived to date after receiving multimodal therapy, including surgery, radiotherapy, chemotherapy, and immunotherapy. Of these, two patients showed no signs of recurrence and progression, and the only patient with recurrence due to also achieved disease control after immunotherapy was re-initiated. To our knowledge, this is the first report of the use of immunotherapy in all treatment phases of patients with SDSC who showed good clinical remission. Only one previous study reported the use of PD-1 inhibitors in patients with SDSC. ⁶⁴ However, immunotherapy was not applied in that study until after radical surgery and adjuvant radiotherapy were completed. ⁶⁴ Therefore, it was not possible to illustrate the clinical control effect of PD-1 inhibitors on the tumors of patients with SDSC by efficacy assessment. Our first patient was previously reported by our colleagues Yang et al., but their report focused mainly on the specificity of pathologic findings. We provide a detailed description of the patient's postoperative treatment regimen and clinical outcomes. ⁶⁵ Owing to the rarity of SDSC, it is difficult to design large-sample rigorous clinical trials to validate the clinical remission and survival benefit of immunotherapy. Furthermore, the limited availability of specimens prevents a comprehensive exploration of the molecular mechanisms driving SDSC and its relationship with immunotherapy. By sharing our treatment experience through a case series, we aim to provide new insights into SDSC management and lay a foundation for future research. To overcome these limitations, we plan to collaborate with other institutions to collect additional samples, focusing on the tumor microenvironment of SDSC and resistance mechanism s to PD-1/PD-L1 inhibitors, ultimately guiding the development of more effective and targeted therapies.

Conclusions

Overall, we report three patients with SDSC who may benefit from anti-PD-1/PD-L1 therapy, thereby providing a new therapeutic option for patients with SDSC. SMARCB1/INI-1 deletion appears to be associated with the efficacy of immunotherapy, although further exploration and validation are necessary.