We report a case of a 1-year-old child with a firm, nonmobile mass in the right thigh, which on CECT was deep-seated and heterogeneously enhanced. Gross evaluation of the excision specimen demonstrated an ill-circumscribed soft tissue mass. Histopathologically, the tumour showed myxoid nodules with primitive mesenchymal spindle-cell areas, and peripheral adipose tissue containing scattered lipoblasts, features suggestive of lipoblastomatosis. The IHC study showed S100 and CD34 positivity in adipocytes, and desmin and myogenin positivity in spindle-cell components. Molecular testing by NGS revealed a COL1A2::PLAG1 gene fusion, thus further establishing the diagnosis. Differentials such as myxoid liposarcoma, fibrous hamartoma of infancy, and infantile fibrosarcoma also harbour distinct genetic drivers, including EWSR1::DDIT3, EGFR exon 20 alterations, and ETV6::NTRK3, respectively.
This case highlights the diagnostic workup of soft-tissue tumours and underscores the importance of integrating histopathology, IHC, and molecular testing. Identification of PLAG1 rearrangement is crucial for confirming the diagnosis. With the growing number of molecularly defined soft-tissue tumours of diagnostic and prognostic significance, incorporating molecular analysis is essential in the evaluation of soft-tissue lesions.
The characteristic histopathological features include a nodular, lobulated architecture with intersecting fibrovascular septa and a spectrum of fat-cell maturation, including primitive mesenchymal cells, lipoblasts, and mature adipocytes, within a myxoid background. Lipomas are benign adipocytic tumours and, although rare in the paediatric age group, should be considered an important differential diagnosis. They are well-circumscribed, encapsulated, and pale yellow, with a homogeneous appearance. Histologically, they are composed of lobules of mature adipose tissue with intervening thin fibrous septa.
A panel of immunohistochemical (IHC) markers, consisting of S100, CD34, and desmin, highlighting various components of the tumour, is essential for diagnosis. However, a definite diagnosis may be challenging at times, owing to tumour heterogeneity, particularly in core needle biopsies. In such scenarios, other paediatric mesenchymal tumours such as lipoma, fibrous hamartoma of infancy, primitive myxoid mesenchymal tumour of infancy, and infantile fibrosarcoma constitute key differentials. Despite the use of an appropriate IHC panel, there remains a risk of misdiagnosis due to clinical and histological overlap and the non-specific immunoprofile of these tumours [4].
PLAG1 (pleomorphic adenoma gene 1) rearrangements, resulting from diverse structural chromosomal alterations involving the 8q11 > q13 (8q12) region, represent the key genetic hallmark of lipoblastoma. With advances in next-generation sequencing (NGS), common fusion partners such as CHCHD7 and COL3A1 have been identified, while rare and novel fusion partners are increasingly being reported [1]. Identification of these molecular genetic alterations is crucial in establishing the definitive diagnosis, particularly in challenging cases, to avoid misdiagnosis and guide treatment decisions.
We report a complex case of lipoblastomatosis in the thigh of a young child, demonstrating a characteristic COL1A2::PLAG1 fusion identified by NGS-based RNA sequencing.
Pathology
The biopsy slides from the outside centre were reviewed, and showed a spindle cell neoplasm with morphology suggestive of fibroblastic/myofibroblastic differentiation, without obvious malignant features; IHC was inconclusive. Subsequently, tissue blocks were sent for an expert soft tissue pathology opinion, which also revealed similar immunomorphology without a definitive diagnosis. In view of the increasing tumour size, the child then underwent wide local excision of the lesion.
Gross evaluation showed an ill-circumscribed, nodular, irregular mass measuring 6.5 x 5.3 cm with adherent soft tissue. The cut surface was solid and nodular, with predominant pale white, firm fibrous areas separating the myxoid nodules and yellow areas in the periphery (Figure 2).
Microscopic evaluation revealed a heterogeneous tumour with lobulated architecture, myxoid nodules, areas of fibrous proliferation, and peripheral adipocytic components. The myxoid areas consisted of plexiform vascular networks and primitive stellate mesenchymal cells (Figure 3). There were wide areas of spindled cells in a fibrous stroma. The periphery of the tumour showed lobules and sheets of adipocytes, including occasional multivacuolated and small signet ring cell lipoblasts (Figure 4). These features represent a spectrum of adipocytic maturation. No areas of marked atypia, necrosis, or increased mitotic activity were noted. The tumour cells were focally seen abutting the inked/cauterised resection margin.
An additional IHC panel was performed in keeping with the earlier marker study. The adipocytic components were immunoreactive for S100 and CD34, whereas the spindle cells showed immunopositivity for desmin and myogenin (focal) (Figure 5). CD34 also highlighted the blood vessels. These tumour cells were negative for ALK, PanTRK, beta-catenin, and MDM2.
With the characteristic histomorphological spectrum of adipocytes at various stages of maturation and corresponding immunoprofile, in conjunction with the clinical and radiological findings of an ill-defined, deep-seated tumour, a final diagnosis of lipoblastomatosis was rendered.
Subsequently, to establish the underlying molecular alteration, the tissue sample was sent for sequencing analysis.
Molecular Analysis
Molecular analysis was performed on the NGS platform (Illumina, Inc., San Diego, California) with a laboratory-developed multi-gene panel covering 108 key genes, including SNVs, indels, and gene fusions. Nucleic acid was extracted from the FFPE tissue block (tumour cellularity ~65%), followed by library preparation using a custom hybrid capture kit. The QC-passed libraries were sequenced to a minimum depth of 250x on the sequencing platform. The variants were annotated using an in-house pipeline, and genomic alterations were prioritised, classified, and reported in accordance with AMP/ASCO/CAP guidelines.
Sequencing analysis revealed a pathogenic variant of COL1A2::PLAG1 fusion with a read depth of 28x. This represents a tier II pathogenic variant of diagnostic significance, with no specific therapeutic implications. No co-occurring mutations were detected. The child remains asymptomatic with no evidence of loco-regional recurrence at 12-month follow-up.
Lipoblastoma is a rare benign soft-tissue tumour of infancy and early childhood. Zhang et al., in their review of 36 cases, report that 78% of cases occurred within 3 years of age, similar to our case, in which the child presented at 1 year of age. Lipoblastomatosis, in contrast, is multicentric, more aggressive, and locally infiltrative with poorly defined borders [8]. These tumours are typically painless and commonly occur in the trunk and extremities; however, unusual sites such as the chest wall, mediastinum, retroperitoneum, and mesentery have been reported [9]. They are usually less than 5 cm in size, but larger lesions up to 25 cm have been reported in the literature [10]. In our case, the lesion measured 5.7 cm and was located in the thigh in a deep soft tissue plane abutting the vessels, making the surgical planes difficult.
Histomorphologically, these tumours exhibit lobulated myxoid nodules with intervening fibrovascular areas and a spectrum of adipocyte cell maturation, ranging from primitive stellate or spindled mesenchymal cells to multivacuolated, small signet-ring cell lipoblasts. The resection specimens show typical zonation, with immature myxoid cells at the periphery and mature adipocytes at the centre. Other variations such as chondroid metaplasia, extramedullary haematopoiesis and multinucleated/floret cells are rarely noted [11]. The diagnosis may be relatively straightforward in an appropriate clinical context. However, a definitive diagnosis by needle biopsy is challenging owing to the tumour's heterogeneous architectural and cellular patterns.
The differential diagnoses when myxoid-predominant areas are encountered include paediatric myxoid liposarcoma and lipoma. Paediatric myxoid liposarcoma is rare in this age group, whereas most of the lipoblastomas are seen in children less than 3 years of age [2]. Lipoma, although rare in children, lacks lipoblasts or primitive mesenchymal cells. The list of possible differentials is broad when a spindle cell (primitive mesenchyme)-predominant area is encountered, including infantile fibrosarcoma, fibrous hamartoma of infancy, inflammatory myofibroblastic tumours (usually seen in the abdomen or chest), and, rarely, primitive myxoid mesenchymal tumour of infancy and ectomesenchymoma.
Although histologically similar, lipoblastoma differs from lipoblastomatosis in its growth pattern. Lipoblastoma is a well-circumscribed, encapsulated tumour, whereas lipoblastomatosis is diffuse, non-encapsulated, and infiltrating, with a higher risk of local recurrence.
Immunohistochemistry plays a pivotal role in narrowing the differential diagnosis; hence, judicious selection of an appropriate panel and careful interpretation of results are crucial. The typical IHC findings in lipoblastoma are as described in our case. We did not have access to PLAG1 by IHC, which is positive in ~80% of cases and shows intense diffuse nuclear staining in adipocytes and spindle cells [12]. PLAG1 expression is a reliable surrogate for PLAG1 rearrangements, and Warren et al. reported a sensitivity of 94% for PLAG1 IHC as compared to FISH / molecular techniques in lipoblastomas [12].
Molecular genetics is crucial for the diagnosis and classification of soft tissue tumours by identifying specific molecular genetic alterations [13]. Thus, it facilitates the distinction between types of soft tissue tumours and makes a definite diagnosis possible when histological overlap exists.
The PLAG1 gene, located on chr8, encodes a zinc finger protein with 2 putative nuclear localization signals, which functions as a transcription factor and regulates cell growth and proliferation. PLAG1 alterations are associated with various cancers and can impact gene expression, including the growth-promoting gene IGF2 [14]. PLAG1 rearrangements involving the chromosomal region 8q12 represent the defining genetic hallmark of lipoblastoma. These rearrangements typically arise from diverse structural chromosomal events, such as minor inversions, insertions or circular chromosomes that lead to promoter swapping and subsequent PLAG1 over-expression.
In lipoblastoma, the most frequently reported PLAG1 fusion partners are CHCHD7::PLAG1 and COL3A1::PLAG1, both of which drive aberrant PLAG1 transcription. Currently, with the increasing use of NGS, a spectrum of fusion partners has been documented, including HAS2, RAD51B, TCEA1, RAB2A, CTDSP2, and ACTA2. In our case, we detected a COL1A2::PLAG1 fusion, accounting for 25% of all PLAG1 fusions in lipoblastoma and representing a tier II pathogenic variant.
Beyond lipoblastoma, PLAG1 rearrangements are implicated in a wide variety of tumours, including pleomorphic adenoma, uterine tumours resembling ovarian sex cord tumours (UTROSCT), and myofibroblastoma of the breast, and are rarely implicated in hepatoblastoma. This underscores the broad relevance and necessitates its documentation [15].
Among paediatric soft-tissue tumours, PLAG1 alterations serve as a valuable diagnostic marker, especially when there is significant immunohistochemical overlap. The key differentials, such as myxoid liposarcoma, fibrous hamartoma of infancy, and infantile fibrosarcoma, also harbour distinct genetic drivers, including EWSR1::DDIT3, EGFR exon20 alterations, and ETV6::NTRK3, respectively. Thus, identification of specific genetic alterations reinforces the diagnosis and helps avoid overtreatment, particularly in paediatric patients [16].
Although FISH is a reliable tool for detecting gene rearrangements, NGS offers broader advantages, such as characterization of fusion partners, identification of cryptic rearrangements, and detection of rare or novel genetic alterations. With declining sequencing costs and increasing accessibility, NGS is poised to become the preferred diagnostic tool, enriching our understanding of tumour biology and refining diagnostic accuracy in soft-tissue pathology.
1) Fritchie K, Wang L, Yin Z, Nakitandwe J, Hedges D, Horvai A, Mora JT, Folpe AL, Bahrami A. Lipoblastomas presenting in older children and adults: analysis of 22 cases with identification of novel PLAG1 fusion partners. Mod Pathol. 2021;34(3):584-91. doi:10.1038/s41379-020-00696-4.
2) Jandali D, Heilingoetter A, Ghai R, Jeffe J, Al-Khudari S. Large parotid gland lipoblastoma in a teenager. Front Pediatr. 2018;6:50.
3) Bourelle S, Viehweger E, Launay F, Quilichini B, Bouvier C, Hagemeijer A, Jouve JL, Bollini G. Lipoblastoma and lipoblastomatosis. J Pediatr Orthop B. 2006;15(5):356-61.
4) Martos-Cabrera L, Sampedro-Ruiz R, Perez-Gonzalez YC, Mentzel T, Llamas-Velasco M. Fibrous hamartoma of infancy: A series of 21 cases and review of the literature. Actas Dermosifiliogr. 2021;112(6):520-7.
5) Choi JH, Ro JY. The recent advances in molecular diagnosis of soft tissue tumours. Int J Mol Sci. 2023;24(6):5934. doi:10.3390/ijms24065934.
6) Chintamani. Soft tissue sarcomas-the pitfalls in diagnosis and management. Indian J Surg Oncol. 2011;2(4):261-4. doi:10.1007/s13193-012-0141-7.
7) Black JO, Al-Ibraheemi A, Arnold MA, Coffin CM, Davis JL, Parham DM, Rudzinski ER, Shenoy A, Surrey LF, Tan SY, Spunt SL. The pathologic diagnosis of pediatric soft tissue tumours in the era of molecular medicine: the Sarcoma Pediatric Pathology Research Interest Group perspective. Arch Pathol Lab Med. 2024;148(1):107-16.
8) Zhang W, Zhang S, Yang Z, Zhang Y, Wang Z. Lipoblastoma in one adult and 35 pediatric patients: Retrospective analysis of 36 cases. Exp Ther Med. 2022;25(1):11. doi:10.3892/etm.2022.11710.
9) Osman B, Kanafani D, Abboud M, Akel S. Pediatric lipoblastomas: a case series. J Pediatr Surg Case Rep. 2023;98:102716. doi:10.1016/j.epsc.2023.102716.
10) Safavi M, Akhlaghi N, Ashjaei B. Large intrathoracic lipoblastoma. Asian Cardiovasc Thorac Ann. 2021;29(8):858-61. doi:10.1177/02184923211038407.
11) Ameloot E, Cordier F, Van Dorpe J, Creytens D. Update of pediatric lipomatous lesions: a clinicopathological, immunohistochemical and molecular overview. J Clin Med. 2022;11(7):1938. doi:10.3390/jcm11071938.
12) Warren M, Tiwari N, Sy S, Raca G, Schmidt RJ, Pawel B. PLAG1 immunohistochemical staining is a surrogate marker for PLAG1 fusions in lipoblastomas. Pediatr Dev Pathol. 2022;25(2):134-40. doi:10.1177/10935266211043366.
13) Chang CC, Shidham VB. Molecular genetics of pediatric soft tissue tumours: clinical application. J Mol Diagn. 2003;5(3):143-54. doi:10.1016/S1525-1578(10)60466-7.
14) D'Occhio MJ, Campanile G, Baruselli PS, Porto Neto LR, Hayes BJ, Snr AC, Fortes MRS. Pleomorphic adenoma gene 1 in reproduction and implication for embryonic survival in cattle: a review. J Anim Sci. 2024;102:skae103. doi:10.1093/jas/skae103.
){kind=link}
){kind=link}
){kind=link}
){kind=link}
){kind=link}