The information below is solely provided to demonstrate the functionalities and features of the RDOD.
Nested Stromal-Epithelial Tumor
Site editors: Viraj Garg BS and Shivani Sud MD
Background
Nested stromal-epithelial tumor (NSET) is a rare tumor of the liver first described in 2001 [1]. The tumor, first referred to as ossifying stromal-epithelial tumor, has gone by several names in the literature including calcifying nested stromal-epithelial tumor, desmoplastic nested spindle cells tumor, and ossifying malignant mixed epithelial and stromal tumor. Though the tumor appears in the World Health Organization’s International Classification of Diseases it lacks well-defined diagnostic criteria, workup, and treatment protocol. In general, the tumor is characterized as non-hepatocytic and non-biliary and occurs predominantly in the right lobe of the liver, though a few cases of left lobe presentation have been reported [2].
Though the histogenesis of the tumor remains largely unknown, it does exhibit some distinguishing histopathological features. NSET is characterized by nests of epithelioid cells surrounded by spindle-shaped stroma cells. In many cases these nests are surrounded by dense myofibroblastic stroma that can be desmoplastic or collagenous in shape [3]. Indeed, it was based on these features and distinct immunohistochemical stains that the tumor was described as a new entity [1]. Several cases in the literature have also reported the presence of osteoid production with calcification and ossification often presenting as key features [2]. It is due to these diverse features that the tumor takes on many different names in the literature. Although histopathologically unique, the rarity of the tumor makes is susceptible to misdiagnosis. Differential diagnosis includes hepatoblastoma, calcified hemangioma, and desmoplastic small round cell tumor, among others. Though misdiagnosis is common, it does not appear to negatively influence treatment or outcome.
To our knowledge, approximately 50 cases on NSET have been reported in the literature with only a handful of cases involving metastasis or recurrence. In addition, a handful of cases in the literature were associated with underlying conditions including Beckwith-Wiedemann Syndrome and Cushing’s Syndrome with no evidence that these underlying conditions complicate the treatment course. Although the presence of Beckwith-Wiedemann Syndrome suggests a potential genetic predisposition, no mutation or genetic abnormality appears reliably across the cases [2].
We analyzed the currently reported cases of NSET collected from the literature through database searches. Case report data was analyzed with the aim of understanding the most common presenting features of the tumor including lab abnormalities, immunohistochemical data, and histopathological data. Due to a lack of current diagnostic criteria and a relatively high rate of misdiagnosis, we hope to use our data to present consolidated diagnostic criteria for use in future cases of NSET. In addition, NSET is a relatively new entity that lacks a formally defined workup and treatment protocol. We aimed to understand the most common presenting features, treatment practices and their associated outcomes. We hope to use the results of this analysis to offer our own standard for the diagnosis and treatment of NSET in order to simplify future management of the disease.
Though the histogenesis of the tumor remains largely unknown, it does exhibit some distinguishing histopathological features. NSET is characterized by nests of epithelioid cells surrounded by spindle-shaped stroma cells. In many cases these nests are surrounded by dense myofibroblastic stroma that can be desmoplastic or collagenous in shape [3]. Indeed, it was based on these features and distinct immunohistochemical stains that the tumor was described as a new entity [1]. Several cases in the literature have also reported the presence of osteoid production with calcification and ossification often presenting as key features [2]. It is due to these diverse features that the tumor takes on many different names in the literature. Although histopathologically unique, the rarity of the tumor makes is susceptible to misdiagnosis. Differential diagnosis includes hepatoblastoma, calcified hemangioma, and desmoplastic small round cell tumor, among others. Though misdiagnosis is common, it does not appear to negatively influence treatment or outcome.
To our knowledge, approximately 50 cases on NSET have been reported in the literature with only a handful of cases involving metastasis or recurrence. In addition, a handful of cases in the literature were associated with underlying conditions including Beckwith-Wiedemann Syndrome and Cushing’s Syndrome with no evidence that these underlying conditions complicate the treatment course. Although the presence of Beckwith-Wiedemann Syndrome suggests a potential genetic predisposition, no mutation or genetic abnormality appears reliably across the cases [2].
We analyzed the currently reported cases of NSET collected from the literature through database searches. Case report data was analyzed with the aim of understanding the most common presenting features of the tumor including lab abnormalities, immunohistochemical data, and histopathological data. Due to a lack of current diagnostic criteria and a relatively high rate of misdiagnosis, we hope to use our data to present consolidated diagnostic criteria for use in future cases of NSET. In addition, NSET is a relatively new entity that lacks a formally defined workup and treatment protocol. We aimed to understand the most common presenting features, treatment practices and their associated outcomes. We hope to use the results of this analysis to offer our own standard for the diagnosis and treatment of NSET in order to simplify future management of the disease.
Cases Published to Date
Approximately 50 cases of NSET have been published in the medical literature to date. We have created a running systematic review table by extracting pertinent clinical information from each of these reports to facilitate clinical and research aims. Tables are organized with the headers below.
nested_stromal_epithelial_tumor_case_listing_table__modified_for_posting_.xlsx |
Cases Published by RDOD
The University of North Carolina at Chapel Hill has notified RDOD of a NSET case under treatment at their hospital. The case will be submitted RDOD as soon as outcome information becomes available.
Methods
English-language case reports and case series published in the literature were collected through searches of PubMed and Google Scholar. Due to the rarity of NSET it has yet to be fully indexed in MedLine. As such, a Google Scholar search was added to supplement the PubMed search to ensure that full body of literature was examined. Keywords for the searches were selected based on the most commonly used terms in the literature [3]. Searches were therefore conducted using the terms “nested-stromal epithelial tumor”, “desmoplastic nested spindle cell tumor”, and “calcifying nested stromal epithelial tumor”.
Prior to data abstraction, all studies were dual screened, based on the PRISMA protocol (Figure 1), by title and abstract for inclusion criteria. Any preclinical studies, cost analysis studies, studies available in abstract only, commentary/editorial articles, and otherwise irrelevant studies were excluded. In addition, studies written in languages other than English were excluded from the data abstraction phase. All studies were screened by two individuals to ensure inclusion of all eligible studies. In total, data was abstracted from 50 cases of NSET.
During data abstraction, both basic article data and individual patient data were abstracted. Article data included publication information such as year and location of publication. Individual
patient data included basic demographics, information regarding disease site, metastases at presentation, presenting symptoms, physical exam findings, lab abnormalities, and gross appearance. Following this, detailed information regarding histopathology, immunohistochemistry, and imaging was recorded.
After collection of basic demographic, presentation, and diagnostic data, in-depth treatment data was abstracted. A basic treatment overview was written to provide a timeline of treatments as well as information regarding post-treatment metastases and recurrences. Data regarding each treatment cycle, if there were multiple, was abstracted including surgical techniques, radiation, and systemic therapies. Also included were any specifics regarding dosages and techniques. The process was repeated for any subsequent treatment courses including for metastasis. In addition, information regarding post-transplant immunosuppression was recorded. Data was also collected with regard to outcome and follow-up, if provided. The cause of death, if appropriate, was also noted as well as time to death after treatment.
After initial data abstraction, the data was further organized based on the most common features present across cases. Many cases, especially those included as part of a case series, included only partial or sparse data. In these cases, any data that was included was abstracted and included in the analysis. To further characterize the current approach to NSET, areas where info was not provided were also documented.
Prior to data abstraction, all studies were dual screened, based on the PRISMA protocol (Figure 1), by title and abstract for inclusion criteria. Any preclinical studies, cost analysis studies, studies available in abstract only, commentary/editorial articles, and otherwise irrelevant studies were excluded. In addition, studies written in languages other than English were excluded from the data abstraction phase. All studies were screened by two individuals to ensure inclusion of all eligible studies. In total, data was abstracted from 50 cases of NSET.
During data abstraction, both basic article data and individual patient data were abstracted. Article data included publication information such as year and location of publication. Individual
patient data included basic demographics, information regarding disease site, metastases at presentation, presenting symptoms, physical exam findings, lab abnormalities, and gross appearance. Following this, detailed information regarding histopathology, immunohistochemistry, and imaging was recorded.
After collection of basic demographic, presentation, and diagnostic data, in-depth treatment data was abstracted. A basic treatment overview was written to provide a timeline of treatments as well as information regarding post-treatment metastases and recurrences. Data regarding each treatment cycle, if there were multiple, was abstracted including surgical techniques, radiation, and systemic therapies. Also included were any specifics regarding dosages and techniques. The process was repeated for any subsequent treatment courses including for metastasis. In addition, information regarding post-transplant immunosuppression was recorded. Data was also collected with regard to outcome and follow-up, if provided. The cause of death, if appropriate, was also noted as well as time to death after treatment.
After initial data abstraction, the data was further organized based on the most common features present across cases. Many cases, especially those included as part of a case series, included only partial or sparse data. In these cases, any data that was included was abstracted and included in the analysis. To further characterize the current approach to NSET, areas where info was not provided were also documented.
Results
Presentation
Cases were analyzed according to presentation. The most common presenting features were abdominal pain/swelling (n = 25; 48%) and incidental presentation (n = 15; 30%). The least common presenting features were lab abnormalities (n = 3; 6%) and weight loss (n = 4; 8%). Cushingoid features (n = 9; 18%) and abdominal mass on physical exam (n=7; 14%) were also relatively rare.
Imaging
Cases were analyzed according to imaging modalities utilized. The most commonly utilized imaging modality, CT (n = 33; 66%), was used to determine whether the mass was heterogeneous (n = 9; 18%), showed calcifications (n = 16; 32%), and showed enhancement (n = 9; 18%). Ultrasound (n = 14; 28%) and MRI (n = 12; 24%) were used but data on their use was limited. MRI was conducted on both the abdomen (n = 11; 92%) and pituitary gland (n = 2; 8%). A PET Scan (n = 8; 16%) was only used as handful of times and only 50% of masses on which it was used were avid.
Pathology
Cases were analyzed for common histological features and immunohistochemical markers (Tables 1 and 2). On microscopy, tumors were characterized by organoid arrangements of epithelioid nests with sheet-like cellular growth. The nests were surrounded by spindle-shape stromal cells and in a minority of cases by myofibroblastic spindle cells that were desmoplastic in nature. Across the tumors there were variable levels of psammomatous calcifications that presented with and without ossification. There was variable involvement of bile ducts and rare occurrences of central necrosis. Vascular and lymphatic invasion were also rare occurrences. All tumors were mitotically active though the extent of the mitotic figures was variable without some tumors showing high levels of activity and several showing very little activity.
Immunohistochemically, epithelioid markers were most consistently expressed across cases. In particular, cytokeratin AE1/AE3 was expressed in 100% of cases and WT-1 in 85% of cases. Presence of these two markers is highly consistent with a diagnosis of NSET. Furthermore, NSET is characterized by a lack of hepatocytic markers. None of the hepatocytic markers tested across tumors came back positive, a finding that is consistent with NSET being a non-hepatocytic tumor. Two stromal markers appear to have strong correlations with NSET. Desmin was negative across all cases and vimentin was positive across all cases. Finally, the neuroendocrine markers showed relatively equal levels of positive and negative expression. Of these markers, chromogranin A was expressed most consistently across cases with a 100% positivity rate. Ki-67 proliferation values were variable with most tumors having low to moderate proliferation. |
Treatment
Surgery is the preferred treatment option for NSET and was used as the first line of treatment in every case for which treatment information was available (n = 46; 92%). More specifically, hepatectomy was the most favored surgical procedure (n = 31; 67%). In some cases, a full transplant (n = 6; 13%) or resection of the tumor (n = 9; 20%) were conducted in lieu of a hepatectomy.
Outcomes
Post-treatment outcomes were widely reported across the cases. Only a handful of cases (n = 5; 10%) failed to report any outcome information. An additional 3 cases (6%) reported some outcome information but failed to provide time points for follow-up. Therefore, outcomes were analyzed based on the cases (n = 47; 94%) for which full outcome information was available.
Kaplan-Maier analysis was conducted to assess overall and progression-free survival in non-metastatic NSET patients (Figures 2 and 3). Overall, the survival rate of NSET is quite high (N = 42; 89.4%) with only 5 patients dying as a result of liver failure, tumor recurrence, or treatment complications (Table 3). In addition, NSET is relatively unlikely to progress or recur. 77% (N = 36) of patients were disease and progression-free at the time of last follow-up (Table 4).
Kaplan-Maier analysis was conducted to assess overall and progression-free survival in non-metastatic NSET patients (Figures 2 and 3). Overall, the survival rate of NSET is quite high (N = 42; 89.4%) with only 5 patients dying as a result of liver failure, tumor recurrence, or treatment complications (Table 3). In addition, NSET is relatively unlikely to progress or recur. 77% (N = 36) of patients were disease and progression-free at the time of last follow-up (Table 4).
Although surgery was the most common treatment for NSET, it appears that quality of surgery is not predictive of outcome. Information regarding quality of the surgery was limited (n = 13; 28%) and was only reported if the margins were tumor-free. Of the cases in which tumor-free margins were reported, one patient had a local recurrence, metastasis, and eventually died from their disease. Another three patients (23%) went on to have a recurrence of the disease and one more patient died from their disease after successful surgery.
Kaplan-Maier analysis was also conducted to assess the efficacy of chemotherapy as a treatment method for non-metastatic NSET. Overall, patients who receive adjuvant chemotherapy have a higher survival rate compared to patients who do not receive chemotherapy. There was a non-statistically significant difference in overall survival between patients who received chemotherapy and those who did not (Figure 4). Nevertheless, all 5 deaths occurred in patients who did not receive chemotherapy and patients who did receive chemotherapy (N = 11) had a 100% survival rate. There was a significant difference in progression-free survival between patients who receive chemotherapy and those who did not (Figure 5). All 11 patients who had local recurrences and/or metastasis never received chemotherapy as part of their initial treatment. Patients who did receive chemotherapy had a 0% progression rate.
Kaplan-Maier analysis was also conducted to assess the efficacy of chemotherapy as a treatment method for non-metastatic NSET. Overall, patients who receive adjuvant chemotherapy have a higher survival rate compared to patients who do not receive chemotherapy. There was a non-statistically significant difference in overall survival between patients who received chemotherapy and those who did not (Figure 4). Nevertheless, all 5 deaths occurred in patients who did not receive chemotherapy and patients who did receive chemotherapy (N = 11) had a 100% survival rate. There was a significant difference in progression-free survival between patients who receive chemotherapy and those who did not (Figure 5). All 11 patients who had local recurrences and/or metastasis never received chemotherapy as part of their initial treatment. Patients who did receive chemotherapy had a 0% progression rate.
Discussion and Future Directions
A systematic review of published nested stromal epithelial tumor cases reports was conducted. English language case-reports and case series were collected through searches on PubMed and Google Scholar. All cases were dual-screened by title and abstract based on PRISMA guidelines for inclusion criteria. Data was abstracted from all included cases. Basic article data as well as individual patient data such as demographics, presentation, diagnosis, and diagnosis were abstracted. Following this, in-depth treatment and outcome data were abstracted for later analysis. Survival of NSET patients was analyzed using the Kaplan-Maier method for generating survival curves. Overall and progression-free survival was analyzed for all non-metastatic cases of NSET where progressions were defined as any local recurrences, metastasis, or deaths that occurred after initial treatment and were related to the tumor. Overall and progression-free survival was also compared for patients who received adjuvant chemotherapy versus those who did not.
To our knowledge, 50 cases of NSET have been reported in the literature in patients of varying ages and health statuses. The tumor appears to be non-aggressive and relatively treatable if caught early enough. Given that the tumor does not present with characteristic genetic mutations, histological and immunohistochemical markers have to be used to confirm diagnosis. Treatment course consists primarily of complete or partial hepatectomy followed in some cases by chemotherapy. Successful treatment appears to prevent recurrence or metastasis of the tumor except in a few rare cases. Even with progression, the tumor is still reasonably treatable with follow-up surgery and chemotherapy.
Although the most common presenting symptom for NSET is abdominal pain and swelling, a large minority of cases presented incidentally. Therefore, it is prudent for clinicians to continue monitoring liver function with blood testing and imaging. Though Cushingoid symptoms were present in a handful of cases, these symptoms do not present reliably across NSET cases. Most commonly, CT was utilized to image tumors to screen for the presence of heterogeneity and calcifications. More important to confirming NSET diagnosis, however, is histological and immunohistochemical analysis. Histologically, NSET presents with characteristic nests of epithelioid cells surrounded by spindle-shaped stromal cells. The presence of these features is a strong indicator of an NSET diagnosis. Furthermore, there exist specific immunohistochemical markers that are strongly associated with an NSET diagnosis. Positive tests for cytokeratin AE1/AE3 and WT-1 are strongly associated with NSET. Positive tests for vimentin and chromogranin A are also characteristic of the tumor.
Treatment course for NSET requires that surgery be conducted. Depending on the extent of liver damage either a complete or partial hepatectomy will be required. In a small number of cases the tumor can be excised without conducting a complete liver transplant. The Kaplan-Maier analysis indicated that the first 5-6 years of disease are the most critical. All deaths and progression reported occurred within the first 5-6 years of disease onset. Patients who made it past this period had a 100% survival rate and 0% progression rate. Although median follow-up time was only 1.37 years, it is clear that NSET can be treated and progression avoided if the tumor is detected as early as possible.
The efficacy of adjuvant chemotherapy was also considered using a Kaplan-Maier analysis. Based on the generated survival curves, there was no statistical difference in overall survival between patients who received adjuvant chemotherapy and those who did not. However, there was a significant difference in progression (p = .046). Patients who were given adjuvant chemotherapy had no progression of disease after initial treatment. Of the patients who were not given adjuvant chemotherapy, 11 recurred and 5 died. Although the data at this point are not randomized, these results indicate that there are potential benefits to administration of adjuvant chemotherapy following surgical resection.
The lack of a standardized format for case reports represented a significant hurdle during the data abstraction phase. Reports varied in the type and depth of information they provided. In particular, there was a significant amount of missing data regarding physical exam findings (n = 36; 72%) and past medical history (n = 34; 68%). In addition, several cases were missing information regarding lab results (n = 15; 30%) and imaging results (n = 13; 26%). 8 cases (16%) reported the presentation of the tumor as incidental but subsequently failed to provide any additional information and 9 cases (18%) failed to report any information regarding immunohistochemical staining. Treatment and outcome data also proved to be sparse in many areas due in part to the fact that case reports cannot be updated. In several reports, the authors indicated they would continue to follow-up with patients but up to date information was not available. 5 (10%) cases failed to report any outcome information and 1 of these cases indicated that the patient was lost to follow-up. Of the few reports that mentioned use of a chemotherapy, 6 (43%) of them failed to specify the type and dosage of chemotherapy utilized. Although surgery was the most favored method for treatment of NSET, the vast majority of cases (n = 27; 54%) failed to report the surgical margins. Of the cases in which surgical margins were reported, all reported the margins as being tumor-free.
Improved aggregation and presentation of clinical data for NSET and indeed any tumor will go a long way in simplifying systematic reviews in the future. Furthermore, consistency is the type and quality of data presented will ensure that all pertinent data across all cases can be captured at the time of data analysis. A more consistent data set would allow for stronger conclusions to be made surrounding the diagnosis and management for rare tumors such as NSET.
One of the limitations of current NSET data is the lack of genetic mutations identified consistently across cases. Genomic analysis in the future will go a long way in improving the identification and management of rare tumors in the future. Whole genome analyses can be utilized to identify tumors that are more likely to be aggressive and therefore need to be managed more aggressively. Furthermore, identification of consistent mutations across specific tumors will present new targets for treating and even preventing the onset of rare disease. Given that NSET is often found incidentally, identification of specific mutations would provide another avenue for early diagnosis and treatment of the disease. Conducting genomic analysis like this enable future linking of clinical data with tissue repositories and molecular diagnostic techniques.
To our knowledge, 50 cases of NSET have been reported in the literature in patients of varying ages and health statuses. The tumor appears to be non-aggressive and relatively treatable if caught early enough. Given that the tumor does not present with characteristic genetic mutations, histological and immunohistochemical markers have to be used to confirm diagnosis. Treatment course consists primarily of complete or partial hepatectomy followed in some cases by chemotherapy. Successful treatment appears to prevent recurrence or metastasis of the tumor except in a few rare cases. Even with progression, the tumor is still reasonably treatable with follow-up surgery and chemotherapy.
Although the most common presenting symptom for NSET is abdominal pain and swelling, a large minority of cases presented incidentally. Therefore, it is prudent for clinicians to continue monitoring liver function with blood testing and imaging. Though Cushingoid symptoms were present in a handful of cases, these symptoms do not present reliably across NSET cases. Most commonly, CT was utilized to image tumors to screen for the presence of heterogeneity and calcifications. More important to confirming NSET diagnosis, however, is histological and immunohistochemical analysis. Histologically, NSET presents with characteristic nests of epithelioid cells surrounded by spindle-shaped stromal cells. The presence of these features is a strong indicator of an NSET diagnosis. Furthermore, there exist specific immunohistochemical markers that are strongly associated with an NSET diagnosis. Positive tests for cytokeratin AE1/AE3 and WT-1 are strongly associated with NSET. Positive tests for vimentin and chromogranin A are also characteristic of the tumor.
Treatment course for NSET requires that surgery be conducted. Depending on the extent of liver damage either a complete or partial hepatectomy will be required. In a small number of cases the tumor can be excised without conducting a complete liver transplant. The Kaplan-Maier analysis indicated that the first 5-6 years of disease are the most critical. All deaths and progression reported occurred within the first 5-6 years of disease onset. Patients who made it past this period had a 100% survival rate and 0% progression rate. Although median follow-up time was only 1.37 years, it is clear that NSET can be treated and progression avoided if the tumor is detected as early as possible.
The efficacy of adjuvant chemotherapy was also considered using a Kaplan-Maier analysis. Based on the generated survival curves, there was no statistical difference in overall survival between patients who received adjuvant chemotherapy and those who did not. However, there was a significant difference in progression (p = .046). Patients who were given adjuvant chemotherapy had no progression of disease after initial treatment. Of the patients who were not given adjuvant chemotherapy, 11 recurred and 5 died. Although the data at this point are not randomized, these results indicate that there are potential benefits to administration of adjuvant chemotherapy following surgical resection.
The lack of a standardized format for case reports represented a significant hurdle during the data abstraction phase. Reports varied in the type and depth of information they provided. In particular, there was a significant amount of missing data regarding physical exam findings (n = 36; 72%) and past medical history (n = 34; 68%). In addition, several cases were missing information regarding lab results (n = 15; 30%) and imaging results (n = 13; 26%). 8 cases (16%) reported the presentation of the tumor as incidental but subsequently failed to provide any additional information and 9 cases (18%) failed to report any information regarding immunohistochemical staining. Treatment and outcome data also proved to be sparse in many areas due in part to the fact that case reports cannot be updated. In several reports, the authors indicated they would continue to follow-up with patients but up to date information was not available. 5 (10%) cases failed to report any outcome information and 1 of these cases indicated that the patient was lost to follow-up. Of the few reports that mentioned use of a chemotherapy, 6 (43%) of them failed to specify the type and dosage of chemotherapy utilized. Although surgery was the most favored method for treatment of NSET, the vast majority of cases (n = 27; 54%) failed to report the surgical margins. Of the cases in which surgical margins were reported, all reported the margins as being tumor-free.
Improved aggregation and presentation of clinical data for NSET and indeed any tumor will go a long way in simplifying systematic reviews in the future. Furthermore, consistency is the type and quality of data presented will ensure that all pertinent data across all cases can be captured at the time of data analysis. A more consistent data set would allow for stronger conclusions to be made surrounding the diagnosis and management for rare tumors such as NSET.
One of the limitations of current NSET data is the lack of genetic mutations identified consistently across cases. Genomic analysis in the future will go a long way in improving the identification and management of rare tumors in the future. Whole genome analyses can be utilized to identify tumors that are more likely to be aggressive and therefore need to be managed more aggressively. Furthermore, identification of consistent mutations across specific tumors will present new targets for treating and even preventing the onset of rare disease. Given that NSET is often found incidentally, identification of specific mutations would provide another avenue for early diagnosis and treatment of the disease. Conducting genomic analysis like this enable future linking of clinical data with tissue repositories and molecular diagnostic techniques.
References
[1] Ishak K, Goodman Z, Stocker JT. Ossifying Stromal-Epithelial Tumor. In: Atlas of Tumor Pathology: Tumors of the Liver and Intrahepatic Bile Ducts. Armed Forces Institute of Technology; 1999.
[2] Benedict M, Zhang X. Calcifying Nested Stromal-Epithelial Tumor of the Liver: An Update and Literature Review. Archives of Pathology & Laboratory Medicine. 2018;143(2):264-268. doi:10.5858/arpa.2017-0346-RS
[3] Geramizadeh B. Nested Stromal-Epithelial Tumor: A Review. Gastrointestinal Tumors. 2019;6(1-2):1-10. doi:10.1159/000496339.
[2] Benedict M, Zhang X. Calcifying Nested Stromal-Epithelial Tumor of the Liver: An Update and Literature Review. Archives of Pathology & Laboratory Medicine. 2018;143(2):264-268. doi:10.5858/arpa.2017-0346-RS
[3] Geramizadeh B. Nested Stromal-Epithelial Tumor: A Review. Gastrointestinal Tumors. 2019;6(1-2):1-10. doi:10.1159/000496339.