SMARCB1/INI1-deficient primary lung carcinoma with hepatic metastasis

SMARCB1-deficient carcinoma is a recently recognised entity in several organs, but has not been previously described arising in non-myoepithelial lung parenchyma. SMARCB1 (also known as INI1 and SNF5) is a universally expressed gene encoding a core component of SWI/SNF chromatin remodelling complexes. SMARCB1 functions as a tumour suppressor in specific cellular contexts and conditional SMARCB1 deletion in mice leads to malignancy with full penetrance, manifesting as rhabdoid tumours or T-cell lymphomas, underlining the gene's biological importance.1Roberts C.W. Leroux M.M. Fleming M.D. et al.Highly penetrant, rapid tumorigenesis through conditional inversion of the tumor suppressor gene Snf5.Cancer Cell. 2002; 2: 415-425Abstract Full Text Full Text PDF PubMed Scopus (253) Google Scholar Complete loss of SMARCB1 expression is a hallmark of rhabdoid tumours and epithelioid sarcomas, is found in some myoepithelial carcinomas, and has rarely been reported in other sarcomas and poorly differentiated carcinomas of the gastrointestinal tract and pancreas.2Parker N.A. Al-Obaidi A. Deutsch J.M. SMARCB1/INI1-deficient tumors of adulthood.F1000Res. 2020; 9: 662Crossref PubMed Scopus (5) Google Scholar SMARCB1-deficient sinonasal carcinoma is a relatively recently described entity associated with SMARCB1 biallelic deletion. Like other SMARCB1-deficient malignancies, these carcinomas behave aggressively and are often refractory to surgery and standard chemoradiotherapy.3Agaimy A. Hartmann A. Antonescu C.R. et al.SMARCB1 (INI-1)-deficient sinonasal carcinoma: a series of 39 cases expanding the morphologic and clinicopathologic spectrum of a recently described entity.Am J Surg Pathol. 2017; 41: 458-471Crossref PubMed Scopus (131) Google Scholar However, emerging evidence that the loss of SMARCB1 confers unique cellular vulnerabilities has spurred the development of novel targeted therapies and highlights the importance of understanding the specific tumour contexts in which pathogenic SMARCB1 mutations occur. Here we describe, to our knowledge, the first report of a primary SMARCB1-deficient lung parenchymal and non-myoepithelial carcinoma, characterised by somatic biallelic SMARCB1 deletion, cytokeratin and p40 expression. A male never smoker in his sixties, with no history of malignancy, presented with a few months of a cough and was found on computed tomography scanning to have a left lower lobe lung lesion suggestive of a primary malignancy. A fluorodeoxyglucose positron emission tomography (PET) scan showed an 86 mm intensely PET avid lung mass consistent with a primary malignancy with ipsilateral hilar and mediastinal nodal and solitary liver metastases. There was no indication of any other metastases, including on brain magnetic resonance imaging. Core biopsy of the liver lesion revealed a malignant neoplasm comprised of sheets and nests of hyperchromatic tumour cells with nuclear moulding, little cytoplasm, conspicuous mitoses and apoptotic bodies (Fig. 1A), overall resembling a small cell carcinoma. On immunohistochemistry (Fig. 1B–N) tumour cells were positive for cytokeratin CAM5.2, p40 in 70% of cells, and negative for TTF1, chromogranin, synaptophysin, INSM1, NSE and CD56, ALK and ROS1. S100 and SMA were also negative (not shown). The Ki-67 index was 60%. No EGFR variant was identified. PD-L1 expression (SP263 clone) was seen in 20% of tumour cells (1+ to 2+ staining). Biopsy of the lung lesion confirmed it to be the same malignancy as the hepatic lesion, with fluorescence in situ hybridisation showing SS18 and EWSR1 were not rearranged. Despite the morphology and immunophenotype being consistent with poorly differentiated squamous cell carcinoma, the lack of smoking history prompted further evaluation using molecular sequencing analysis of tumour cells by Trusight Oncology 500 assay (Illumina, USA), which showed biallelic SMARCB1 deletion (Fig. 1O). Coincident BCR loss was also identified, and is in keeping with the close proximity of BCR and INI1 loci on chromosome 22 and with previous reports of co-deletion of these genes in rhabdoid tumours.4Perry A. Fuller C.E. Judkins A.R. et al.INI1 expression is retained in composite rhabdoid tumors, including rhabdoid meningiomas.Mod Pathol. 2005; 18: 951-958Crossref PubMed Scopus (131) Google Scholar Loss of SMARCB1 (INI1) protein expression in tumour cells was confirmed using immunohistochemistry. The patient was diagnosed with SMARCB1-deficient carcinoma. The TSO500 assay also found low tumour mutation burden and microsatellite stable status. The patient was staged as having cT4N2M1b disease and received immunotherapy with ipilimumab and nivolumab combined with carboplatin and paclitaxel. He had an initial partial metabolic response to this combination, although subsequently had disease progression when treated with immunotherapy alone. He was rechallenged with gemcitabine/carboplatin and continues to derive benefit from this combination (Fig. 2). SMARCB1-deficient sinonasal carcinoma is a recognised diagnostic entity, although by definition is of sinonasal origin. In a comprehensive analysis, Agaimy et al. showed SMARCB1-deficient sinonasal carcinomas generally have either basaloid or rhabdoid/plasmacytoid morphology, almost invariably express cytokeratins, variably express p63, synaptophysin and chromogranin and uncommonly express p16.3Agaimy A. Hartmann A. Antonescu C.R. et al.SMARCB1 (INI-1)-deficient sinonasal carcinoma: a series of 39 cases expanding the morphologic and clinicopathologic spectrum of a recently described entity.Am J Surg Pathol. 2017; 41: 458-471Crossref PubMed Scopus (131) Google Scholar Epstein–Barr virus (EBV) and human papilloma virus (HPV) were negative. In our case, the morphology was small cell or perhaps basaloid, although with no specific evidence of squamous differentiation on H&E as with basaloid lung squamous cell carcinoma. Our case showed no rhabdoid or plasmacytoid features. There was insufficient tissue to test for p16, HPV and EBV. Regardless, there was clear overlap with our case and SMARCB1-deficient sinonasal carcinoma in terms of cytokeratin positivity and squamous marker expression (p40 or p63). With the recent appreciation that over 20% of tumours harbour mutations in genes encoding SWI/SNF components, understanding the diagnostic, prognostic and therapeutic significance of these mutations has become a high priority.5Centore R.C. Sandoval G.J. Soares L.M.M. et al.Mammalian SWI/SNF chromatin remodeling complexes: emerging mechanisms and therapeutic strategies.Trends Genet. 2020; 36: 936-950Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar SWI/SNF proteins form distinct subcomplexes comprised of a core ATPase (either SMARCA4 or SMARCA2) plus up to 12 additional subunits. SMARCB1 is a core component of the two principal mammalian SWI/SNF complexes. SWI/SNF complexes mobilise nucleosomes to regulate transcription factor binding and thus are fundamental to the normal cellular regulation of gene expression. Deletion or loss of function mutations in these components drives the formation of aberrant SWI/SNF complexes and ensuing transcriptional dysregulation. The cellular consequences of SWI/SNF disruption are highly dependent on the cell context in which the mutations arise. Indeed, unusually for a tumour suppressor gene, in most cancer cell lines SMARCB1 is essential for survival. However, in specific cellular contexts its loss is a potent driver of oncogenesis, accounting for the highly select group of malignancies characterised by SMARCB1 deficiency. Thus, the identification of biallelic SMARCB1 deletion in a primary lung carcinoma is likely to represent a tumour defining driver mutation. Primary lung squamous cell carcinoma is almost universally associated with smoking, therefore the never smoking history and lack of additional mutations further supports a distinct aetiology and pathogenesis for this tumour. Inactivating mutations in SWI/SNF family proteins are relatively frequent in non-small cell lung carcinoma; however, these most commonly involve SMARCA4 and are typically seen in smoking related adenocarcinoma in conjunction with other driver mutations such as KRAS, TP53 and STK11.6Sesboue C. Le Loarer F. SWI/SNF-deficient thoraco-pulmonary neoplasms.Semin Diagn Pathol. 2021; 38: 183-194Crossref PubMed Scopus (8) Google Scholar This case of a p40 positive lung carcinoma with SMARCB1 deletion is a distinct and more unusual scenario. Interestingly, an aggressive pleural SMARCB1-deficient p40 positive carcinoma in a young non-smoker has previously been reported.7Yoshida K. Fujiwara Y. Goto Y. et al.The first case of SMARCB1 (INI1) - deficient squamous cell carcinoma of the pleura: a case report.BMC Cancer. 2018; 18: 398Crossref PubMed Scopus (6) Google Scholar Examining the AACR Project GENIE registry (Version 10),8The AACR Project GENIE ConsortiumAACR Project GENIE: powering precision medicine through an international consortium.Cancer Discov. 2017; 7: 818-831Crossref PubMed Scopus (711) Google Scholar, 9Cerami E. Gao J. Dogrusoz U. et al.The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data.Cancer Discov. 2012; 2: 401-404Crossref PubMed Scopus (9151) Google Scholar, 10Gao J. Aksoy B.A. Dogrusoz U. et al.Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal.Sci Signal. 2013; 6: l1Crossref Scopus (8474) Google Scholar among 3,519 tumours designated as squamous or adenosquamous carcinomas, we identified two with homozygous SMARCB1 deletion, one sinonasal carcinoma and one diagnosed as primary lung squamous cell carcinoma. Amongst all lung cancers (10,974 patients), there were three additional cases with homozygous SMARCB1 deletion; one diagnosed as small cell lung carcinoma, one neuroendocrine tumour and one adenocarcinoma. Therefore, although the GENIE registry comprises data from targeted panels by next generation sequencing which may not be a sensitive method for detecting gene deletion, SMARCB1 deletions appear to be rare in primary lung cancer, and the morphological and immunophenotypic resemblance to SMARCB1-deficient carcinomas of the sinonasal tract and occurrence in never smokers, suggests that pulmonary SMARCB1-deficient carcinoma may represent a distinct diagnostic entity. SWI/SNF mutations are typically associated with an aggressive clinical course but also predict specific therapeutic sensitivities. The EZH2 inhibitor tazemetostat has recently been FDA approved for the treatment of SMARCB1-deficient epithelioid sarcoma, representing the first molecularly targeted therapy for SWI/SNF-deficient malignancies.11Mittal P. Roberts C.W.M. The SWI/SNF complex in cancer - biology, biomarkers and therapy.Nat Rev Clin Oncol. 2020; 17: 435-448Crossref PubMed Scopus (145) Google Scholar Given the large prevalence of non-small cell lung cancers even a rare association is potentially important in terms of absolute numbers, particularly given the poor response to conventional therapy and the potential for targeted therapy in the precision medicine era. The response to chemoimmunotherapy in this case is noteworthy given low PD-L1 expression and low tumour mutation burden are generally associated with decreased immunotherapy responsiveness. While immune checkpoint inhibitors have proved successful in tumours with deregulated SWI/SNF signalling, including in a SMARCB1-deficient model of paediatric rhabdoid tumours, which are characterised by a very low tumour mutation burden,12Leruste A. Tosello J. Ramos R.N. et al.Clonally expanded T cells reveal immunogenicity of rhabdoid tumors.Cancer Cell. 2019; 36: 597-612Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar the predominant determinant of response in this case may have been the chemotherapy component. SMARCB1-deficient malignancies have diverse morphological appearances and immunophenotypes. We suggest that SMARCB1-deficient carcinoma be considered in cases of poorly differentiated lung carcinoma, particularly where there is disconnect between tumour morphology and immunophenotype or clinical risk factors such as a lack of smoking history. In such cases, when available, SMARCB1 (INI1) immunohistochemistry offers an easy way to screen for SMARCB1 deficiency. Awareness and recognition of such cases may help avoid ineffective treatments, sharpen focus on targeted therapeutic development, and expand our understanding of the role and extent of SMARCB1 loss as an oncogenic driver. The authors acknowledge the American Association for Cancer Research and its financial and material support in the development of the AACR Project GENIE registry, as well as members of the consortium for their commitment to data sharing. Interpretations are the responsibility of study authors. Marian L. Burr has received funding from a Cancer Research UK Clinician Scientist Fellowship (C53779/A20097) and NHMRC grants 1164054 and 1196598. The authors state that there are no conflicts of interest to disclose.