Abstract

Gastroblastoma is a rare biphasic epithelial-mesenchymal neoplasm of the stomach with fewer than 20 tumors reported in literature. 1 Most often it is located in the antrum of the stomach, centered in the muscularis propria of the gastric wall. 1 It also has been reported to involve the serosa, and may metastasize to lymph nodes, peritoneum, and other organs.1,2
The morphology of gastroblastoma is characterized by a variation of spindle and epithelial components.1,2 The epithelial component consists of cords of neoplastic cells with indistinct borders, vesicular nuclei, small nuclei, and occasionally eosinophilic secretions. 1 This component stains positive for AE1/AE3, CAM 5.2, keratin 7, keratin 18, CD10, and CD56. 2 The mitotic rate of gastroblastoma is relatively low. 2 Grossly it is often a solid or cystic mass arising from the gastric wall, sometimes with associated hemorrhage and/or necrosis. 2 It has been found to harbor 2 fusion genes, MALAT1::GLI1 and PTCH1::GLI2. 2
The clinical features of gastroblastoma are nonspecific and variable and it can often be misdiagnosed at the time of presentation. 1 The treatment often consists of resection of the primary tumor. 2 There is one report of neoadjuvant chemotherapy that was given to a patient, where the tumor was misdiagnosed as an adenocarcinoma, and it showed no response to therapy. 3
We previously reported a tumor in a 28-year-old man who presented after a motor vehicle collision and was found to have an incidental gastric lesion. 3 This was initially misdiagnosed as adenocarcinoma, and he underwent 6 weeks of chemotherapy with no response. Following this, he had a positron emission tomography (PET) scan which revealed metastatic disease and therefore underwent operative resection of the tumor through partial gastrectomy in order to obtain a tissue diagnosis. At the time of surgery, he was noted to have peritoneal studding, liver metastases and drop metastases to the pelvis. Grossly, the specimen consisted of a 5.5-centimeter (cm) portion of the distal stomach with a transmural, partially circumscribed mass with cystic hemorrhagic focus, measuring 3.8 × 3.3 × 2.5 cm, as well as a few enlarged and firm lymph nodes in perigastric tissue. Microscopically, the tumor arose within the gastric wall and involved the submucosa and muscularis mucosae. It had biphasic morphology with mesenchymal and epithelial components. The mesenchymal component was comprised of sheets of spindle cells while the epithelial component was comprised of epithelial cells arranged in clusters forming gland-like structures. It stained positive for CD56, neuro-specific enolase, and CAM 5.2. Mitotic index ranged from 0 to 30 mitotic figures per 50 high-power fields. Metastatic disease was identified in 1 of 4 perigastric lymph nodes. Resection margins were negative. Histological sections can be seen in Figure 1. Immediately following his resection, he was offered further genomic testing and treatment given his metastases, however he declined.

(A) Gastroblastoma arising within the gastric wall and involving the submucosa and muscularis mucosae, sparing the surface epithelium (Hematoxylin and Eosin [H&E], 4×, 300μm). (B) Biphasic morphology with mesenchymal (arrowhead) and epithelial components forming glands with inspissated material (arrow) (10×, scale bar 125μm). (C) The neoplastic cells are cytologically bland. A sheet of spindle cells (left) comprises the mesenchymal component. The epithelial cells (right) are arranged in clusters and form gland-like structures (H&E, 20×, 45μm). (D) Lymph node metastasis (H&E, 4×, scale bar 300μm).
We aim to provide an update on the literature in terms of long-term survival of patients with metastatic gastroblastoma. In order to do this, we provide an update on our patient's status. Data was collected regarding his symptoms, surveillance, further treatments, and quality of life. We performed genetic testing to assess for hereditary cancer syndromes. We also performed next-generation sequencing-based targeted genomic profiling on his tissue blocks to examine for targetable mutations to inform further treatment options for him and future patients. The test utilized was the sarcoma targeted gene fusion panel which examined 138 genes (Appendix 1). The GLI1 gene, associated with gastroblastoma, is targeted in this panel, however, it does not include the MALAT1, GLI2, or PTCH1 genes which are also commonly associated. This did not identify any mutations in the genes tested. He also underwent additional testing for hereditary cancer syndromes through use of a gastric cancer panel. This panel tested for mutations in the following genes APC, BMPR1A, CDH1, CTNNA1, EPCAM (deletion/duplication only), KIT, MLH1, MSH2, MSH6, NF1, PDGFRA, PMS2, SDHA, SDHB, SDHC, SDHD, SMAD4, STK11, and TP53. No mutations were identified, reducing the likelihood of a hereditary cancer syndrome.
Our patient continues to follow with an oncologist and has been undergoing surveillance PET scans every 3 years which has shown persistent, however stable, fluorodeoxyglucose (FDG)-avid lesions in the right hepatic lobe, right kidney, and retro-vesicular space in the pelvis. He did have one PET scan showing uptake at the gastroesophageal junction and an esophagogastroduodenoscopy was performed. No evidence of tumor recurrence was seen. He remains asymptomatic from his disease 11 years Post-resection and lives with a good quality of life without functional limitations.
This report highlights the indolent nature of gastroblastoma, even in the untreated metastatic form. There have been no other reports describing this length of survival in cases of metastatic gastroblastoma. We also describe the use of hereditary cancer screening and next-generation sequencing-based targeted genomic profiling for gastroblastoma. While next-generation sequencing did not yield any targetable mutations for our patient, this panel did not test for the mutations commonly seen in gastroblastoma. Further testing that includes these genes is warranted. Its use should still be considered for soft tissue tumors as its genomic profile continues to expand.
Footnotes
Ethical Considerations
University of Toledo Institutional Review Board—301706.
Consent to Participate
IRB approval and patient consent was obtained. Patient consent for publication was obtained as well.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Appendix 1. Sarcoma Targeted Gene Panel.
ACTB, AHRR, ALK, ASPSCR1, ATF1, ATIC, BCOR, BRD3, BRD4, CAMTA1, CARS, CCNB3, CDH11, CDX1, CD63, CEP128, CIC, CITED2, CLTC, CNBP, COL1A1, COL1A2, COL3A1, COL6A3, CREB1, CREB3L1, CREB3L2, CSF1, CXorf67, C11orf95, DDIT3, DUX4, DVL2, EML4, EPC1, EP400, ERG, ETV1, ETV4, ETV6, EWSR1, FEV, FGFR1, FLI1, FN1, FOSB, FOXO1, FOXO4, FUS, GLI1, HAS2, HEY1, HMGA2, IRF2BP2, JAZF1, KIRREL, KLF17, LAMTOR1, LPP, MAML3, MBTD, MEAF6, MED12, MIR143HG, MKL2, MYH9, NAB2, NCOA1, NCOA2, NFATC2, NFIB, NOTCH1, NOTCH2, NR4A3, NTRK1, NTRK3, NUMA1, NUTM1, NUTM2B, OMD, OPHN1, PATZ1, PAX3, PAX7, PBX1, PBX3, PDGFB, PDPN, PHF1, PLAG1, PLPP3, POU5F1, PPFIBP1, PRDM10, PRKCA, PRKCB, PRKCD, RAB2A,RAD51B, RANBP2, RNF213, RRAGB, SEC31A, SERPINE1, SETBP1, SFMBT1, SMARCA5, SP3, SQSTM1, SRF, SRSF3, SSX1, SSX2, SSX4, SS18, SS18L1, STAT6, SUZ12, S100A10, TAF15, TCF12, TEAD1, TFE3, TFG, THRAP3, TPM3, TPM4, TPR, USP6, VCL, VGLL2, WT1, WWTR1, YAP1, YWHAE, ZC3H7B, ZFP36, and ZNF444.
