Discussion :
As PPB is the archetype of DICER1- associated neoplasms, it is not surprising that several of the extrapulmonary DICER1 mutated sarcomas share aspects of its morphologic template from a cystic to solid macroscopic appearance, to one or more of the histologic patterns. The basic collage consists of a variable intermixture of rhabdomyosarcoma usually embryonal type, islands and nests of blastema, primitive spindle cell pattern with presence of cambium layer, nodules of fetal to sarcomatous cartilage and anaplastic tumor cells. The various sarcomatous patterns are heterogeneously represented, so that blastema is the dominant pattern in one tumor and spindle cells sarcoma in another. This similarity to PPB was noted by Schultz et al and they proposed the designation ‘pleuropulmonary blastoma-like peritoneal sarcoma’. 8
Warren et al noted in their literature review that mostDICER1- associated sarcomas had a component of undifferentiated small round blue cells with rhabdomyoblastic differentiation (in 64/71 cases; 90.1%), poorly differentiated spindle cell component with anaplasia (45/50; 90%) and chondroid/myxoid differentiation (39/45; 86.7%). This similarity was seen in spite of the different nomenclature used for the reported cases and the varied sites of presentation. Though this was not addressed, the morphologic descriptions have a striking resemblance to PPB. Rhabdomyoblastic differentiation was the most common feature in all these cases, (including in the 3 cases reported by them), while the other features were present in variable frequency. Appreciating the commonality in morphology, they proposed a unified terminology for these tumours – ‘DICER1 -associated sarcoma’, which was further supported by McCluggage et al. 7,10
In our 3 cases, similarities with PPB type III include primitive spindle cell pattern, cambium-like arrangement of tumour cells, nodules of fetal to sarcomatous cartilage and a variable population of anaplastic tumor cells. Chondroid differentiation and anaplastic component was lacking in one case. small focus of immature cartilage was identified in only one out of 11 sections from tumour. Although chondroid differentiation may be focal and could be missed in a biopsy, however, in our experience, when present it is quite specific for DICER1 sarcoma and should prompt testing for DICER1 pathogenic variation. None of the tumors however had a neoplastic epithelial component and rhabdomyoblastic differentiation was the only constant feature in all 3 cases. However, this rhabdomyoblastic differentiation was somewhat patchy in all 3 cases, both morphologically and on immunohistochemical staining (with desmin, myogenin and myoD1), which raised our suspicion and helped avoid misclassification as typical rhabdomyosarcoma.
Another morphologic variation not described in the Warren review, is the rare multi-patterned DICER1 -associated tumour with primitive neuroepithelial elements, designated as ‘pre-sacral malignant teratoid neoplasm in association with pathogenic DICER1 variation’.9 We also noted neuropil with possible focal neuroblastic-like component in Case 3, but in all these tumours and ours, the predominant component was ERMS. Pancaldi et al reported aDICER1 associated PNET with EWSR1 rearrangement which did not show any ERMS component. 11 A unique observation in our case was the positivity of SALL4 and synaptophysin in the recurrent high grade sarcoma, which was negative for myoid markers. Warren and McCluggage have separately reported 2 ovarian tumours, where the small spindle cells were SALL4 positive, but dually positive for myoid markers (desmin and myogenin). 7,12
The DICER1 -associated sarcomas seem to have a predilection for the female genital tract and brain. 7,8,12,13 However, rare cases have been reported in males as well, including the para-testicular region. 14
The PPB-like morphologic spectrum ranging from benign to overtly malignant sarcoma, is also noted in DICER1 associated renal cystic nephromas and anaplastic sarcoma of kidney, both of which also show rhabdomyoblastic differentiation. 7,15
Like in cystic nephromas and type 1 PPB, the spindle cell morphology in some DICER1 associated tumours can be very bland and hypocellular in some tumours and can be misinterpreted on a small biopsy as fibrotic tissue (as with the recurrent vaginal tumour case 3), even though the cells express RMS markers on IHC. This has also been highlighted by Yoon et al. 16 Routine use of desmin to evaluate genital tract spindle cell lesions in young females, may help correctly identify these tumours.
The association of rhabdomyosarcomas of the female genital tract withDICER1 has been reported variably in literature, but it is not clear yet if identification of any ERMS element should prompt consideration of a DICER1 -related neoplasm and justify testing, especially in the absence of immature cartilage or primitive mesenchyme. This question is most pertinent for children where ERMS is known to be one of the commonest childhood sarcomas. While Apellaniz-Ruiz in their review showed that nearly all gynecologic ERMSs (except vaginal tumours) were DICER1 -associated, Doros et al detected DICER1mutation in only 3.8% (2/52) cases of sporadic ERMS and the Children’s oncology group cohort of pediatric rhabdomyosarcomas showedDICER1 germline mutation in 4.4%. 10,17-20
DICER1 -associated central nervous system sarcomas also have a high association with rhabdomyosarcomatous differentiation.8,21,22 Our CNS case mimicked mesenchymal chondrosarcoma, though it expressed ERMS markers.
We found the presence of gain of chromosome 8 in cases 1 and 2 (FISH was attempted in case 3, but failed quality measures). Review of literature suggests a very high incidence of gain of chromosome 8 with PPB.23 Interestingly, association of gain of chromosome 8 is also reported in a high proportion of cases of ERMS and some cases of mesenchymal chondrosarcomas. 24,25 This molecular association has not yet been well explored in DICER1 -related sarcomas, needs further investigation. A recent landmark paper of a DNA methylation based sarcoma classifier found that DICER1 mutated sarcomas cluster independent (methylation class sarcoma RMS-like) from conventional embryonal and MYOD1 mutated rhabdomyosarcomas and may represent a distinct group of tumours. This distinction is further sharpened as these tumours do not respond to standard RMS protocol chemotherapy. All the DICER1 mutated sarcomas in their cohort showed gain of chromosome 8 as well. 26
Treatment guidelines are still not standard for these tumours, though most reports have used an anthracycline and alkylator based protocol with varied response. 8 These tumours seem to benefit from initially a more intensive chemotherapy backbone followed by a prolonged maintenance chemotherapy strategy, in spite of the presence of RMS differentiation, thus making the accurate diagnosis of DICER1mutated sarcoma, vital. All our patients had aggressive disease but responded to the alternating VDC/PEI backbone followed by maintenance chemotherapy. The decision for VDC/PEI for the CNS case was based on a multidisciplinary tumor board review with expert consensus from around the world and then it became our standard for the subsequent tumors in view of the excellent responses observed serially. Satisfactory local control with surgery and/or radiotherapy seems imperative for maintaining remission. Prognosis can be favorable, provided the tumours are detected early. 1,2,27
The risk of developing a tumour associated with germline DICER1variation is greatest in early childhood (1st decade) and decreases with increasing age and surveillance protocols are developed accordingly. Correct and early diagnosis of DICER1syndrome is imperative to ensure appropriate surveillance protocols are followed in these patients for improving survival. Apart from macrocephaly in some individuals, this cancer predisposition syndrome is typically not associated with specific phenotypic characteristics, making clinical identification difficult. Hence histologic identification is the key and the onus rests on the pathologist to correctly identify the morphologic clues to guide testing forDICER1 mutation, especially in resource constrained settings. The presence of rhabdomyoblastic differentiation along with primitive blastemal and mesenchymal component, with or without chondroid differentiation are the key features which should raise suspicion, especially in pediatric sarcomas of the female genitourinary tract or brain.
Disclosure / Conflict of interest : The authors D. Ashley Hill and KrisAnn Schultz are supported by National Institutes of Health grant NCI R01CA143167 and NCI R37CA24494. The International PPB/DICER1Registry gratefully acknowledges philanthropic support from the Pine Tree Apple Classic Fund and Children’s Minnesota Foundation.
The other authors have no conflict of interest and nothing to disclose.
Ethical approval : Institutional review board approval ((EC/WV/TMC/41/20) has been taken for this study.
Informed consent  - Informed consent was taken from patients’ parents before this study
Funding – None. Testing for DICER1 was supported by the DICER1/ Pleuro-Pulmonary Blastoma society. Courier charges were borne by Tata Medical Center.